Busbar benchmark measurements
Operated by the Busbar team; Busbar is also an entrant. Gateway overhead against a local mock is not end-to-end model latency. Methodology and limitations.
Rust · recorded build: getbusbar/busbar:1.5.5. Measured: .
JSON · Markdown · CSV · Project source · Interactive dashboard
Each row below names its protocol pair. A representative dashboard cell may differ between gateways; compare an identical pair before interpreting a ranking. Unavailable metrics retain their reason; below-resolution readings are not exact zeroes.
Published measurements
| Ingress → upstream | Metric | Value | State / reason | p99 bound (ms) |
|---|---|---|---|---|
| Cohere → Anthropic | Added latency p50 (µs) | 99 | Valid · n=49,012 gateway / 264,547 direct | — |
| Cohere → Anthropic | Added latency p99 (µs) | 111 | Valid · n=49,012 gateway / 264,547 direct | — |
| Cohere → Anthropic | Gateway p99 at concurrency 1 (µs) | 139 | Valid · n=49,012 | — |
| Cohere → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=264,547 | — |
| Cohere → Anthropic | CPU per request (µs) | 88.93 | Valid | — |
| Cohere → Anthropic | Requests per CPU-second | 11,245.2 | Valid | — |
| Cohere → Anthropic | Cost window concurrency | 8 | Valid | — |
| Cohere → Anthropic | Cost window core utilisation | 0.967 | Valid | — |
| Cohere → Anthropic | Cost window successful requests | 261,900 | Valid | — |
| Cohere → Anthropic | Cost window throughput (req/s) | 43,646 | Valid | — |
| Cohere → Anthropic | Cost window threads | 10 | Valid | — |
| Cohere → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → Anthropic | Cost window major page faults | 0 | Valid | — |
| Cohere → Anthropic | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → Anthropic | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → Anthropic | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → Anthropic | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → Anthropic | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → Anthropic | Streams sustained | — | Untestable | — |
| Cohere → Anthropic | Idle RSS (MiB) | 7.43 | Valid | — |
| Cohere → Anthropic | Steady-state RSS (MiB) | 24.29 | Valid | — |
| Cohere → Anthropic | Recovered RSS (MiB) | 16.4 | Valid | — |
| Cohere → Anthropic | Peak RSS (MiB) | 24.69 | Valid | — |
| Cohere → Anthropic | Peak RSS high-water mark (MiB) | 24.31 | Valid | — |
| Cohere → Anthropic | Time to plateau (s) | — | Not measured | — |
| Cohere → Anthropic | Memory growth (MiB/min) | 0.523 | Valid | — |
| Cohere → Anthropic | Memory load duration (s) | 300 | Valid | — |
| Cohere → Anthropic | Memory shape under load | 1 | Valid | — |
| Cohere → Anthropic | Idle memory shape | 0 | Valid | — |
| Cohere → Anthropic | Idle memory static | 0 | Valid | — |
| Cohere → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → Anthropic | Throughput (req/s) | 46,645 | concurrency 16; n=279,903 | 1 ms |
| Cohere → Anthropic | Throughput (req/s) | 47,855 | concurrency 128; n=287,174 | 5 ms |
| Cohere → Anthropic | Throughput (req/s) | 47,855 | concurrency 128; n=287,174 | 10 ms |
| Cohere → Anthropic | Throughput (req/s) | 47,855 | concurrency 128; n=287,174 | 50 ms |
| Cohere → Anthropic | Throughput (req/s) | 47,855 | concurrency 128; n=287,174 | 100 ms |
| Cohere → Anthropic | Throughput (req/s) | 47,855 | concurrency 128; n=287,174 | Unbounded |
| OpenAI → Anthropic | Added latency p50 (µs) | 101 | Valid · n=47,933 gateway / 264,624 direct | — |
| OpenAI → Anthropic | Added latency p99 (µs) | 115 | Valid · n=47,933 gateway / 264,624 direct | — |
| OpenAI → Anthropic | Gateway p99 at concurrency 1 (µs) | 143 | Valid · n=47,933 | — |
| OpenAI → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=264,624 | — |
| OpenAI → Anthropic | CPU per request (µs) | 89.26 | Valid | — |
| OpenAI → Anthropic | Requests per CPU-second | 11,203.8 | Valid | — |
| OpenAI → Anthropic | Cost window concurrency | 8 | Valid | — |
| OpenAI → Anthropic | Cost window core utilisation | 0.963 | Valid | — |
| OpenAI → Anthropic | Cost window successful requests | 259,816 | Valid | — |
| OpenAI → Anthropic | Cost window throughput (req/s) | 43,297 | Valid | — |
| OpenAI → Anthropic | Cost window threads | 10 | Valid | — |
| OpenAI → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → Anthropic | Cost window major page faults | 0 | Valid | — |
| OpenAI → Anthropic | Added time to first event p50 (µs) | 154 | Valid · n=100 gateway / 100 direct | — |
| OpenAI → Anthropic | Added time to first event p99 (µs) | 248 | Valid · n=100 gateway / 100 direct | — |
| OpenAI → Anthropic | Added inter-frame gap p50 (µs) | 1 | Valid · n=63 gateway / 63 direct | — |
| OpenAI → Anthropic | Added inter-frame gap p99 (µs) | — | Below resolution · n=63 gateway / 63 direct | — |
| OpenAI → Anthropic | Streams sustained (frames/s) | 41,099.6 | Valid | — |
| OpenAI → Anthropic | Streams sustained | 1,536 | Valid | — |
| OpenAI → Anthropic | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI → Anthropic | Steady-state RSS (MiB) | 20.11 | Valid | — |
| OpenAI → Anthropic | Recovered RSS (MiB) | 15.37 | Valid | — |
| OpenAI → Anthropic | Peak RSS (MiB) | 20.68 | Valid | — |
| OpenAI → Anthropic | Peak RSS high-water mark (MiB) | 20.47 | Valid | — |
| OpenAI → Anthropic | Time to plateau (s) | — | Not measured | — |
| OpenAI → Anthropic | Memory growth (MiB/min) | -0.144 | Valid | — |
| OpenAI → Anthropic | Memory load duration (s) | 300 | Valid | — |
| OpenAI → Anthropic | Memory shape under load | 0 | Valid | — |
| OpenAI → Anthropic | Idle memory shape | 0 | Valid | — |
| OpenAI → Anthropic | Idle memory static | 0 | Valid | — |
| OpenAI → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → Anthropic | Throughput (req/s) | 46,006 | concurrency 16; n=276,061 | 1 ms |
| OpenAI → Anthropic | Throughput (req/s) | 47,478 | concurrency 128; n=284,911 | 5 ms |
| OpenAI → Anthropic | Throughput (req/s) | 47,478 | concurrency 128; n=284,911 | 10 ms |
| OpenAI → Anthropic | Throughput (req/s) | 47,478 | concurrency 128; n=284,911 | 50 ms |
| OpenAI → Anthropic | Throughput (req/s) | 47,478 | concurrency 128; n=284,911 | 100 ms |
| OpenAI → Anthropic | Throughput (req/s) | 47,478 | concurrency 128; n=284,911 | Unbounded |
| OpenAI Responses → Anthropic | Added latency p50 (µs) | 150 | Valid · n=34,460 gateway / 264,688 direct | — |
| OpenAI Responses → Anthropic | Added latency p99 (µs) | 166 | Valid · n=34,460 gateway / 264,688 direct | — |
| OpenAI Responses → Anthropic | Gateway p99 at concurrency 1 (µs) | 193 | Valid · n=34,460 | — |
| OpenAI Responses → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=264,688 | — |
| OpenAI Responses → Anthropic | CPU per request (µs) | 136.4 | Valid | — |
| OpenAI Responses → Anthropic | Requests per CPU-second | 7,329.7 | Valid | — |
| OpenAI Responses → Anthropic | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → Anthropic | Cost window core utilisation | 0.98 | Valid | — |
| OpenAI Responses → Anthropic | Cost window successful requests | 173,055 | Valid | — |
| OpenAI Responses → Anthropic | Cost window throughput (req/s) | 28,839 | Valid | — |
| OpenAI Responses → Anthropic | Cost window threads | 10 | Valid | — |
| OpenAI Responses → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → Anthropic | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → Anthropic | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → Anthropic | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → Anthropic | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → Anthropic | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → Anthropic | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → Anthropic | Streams sustained | — | Untestable | — |
| OpenAI Responses → Anthropic | Idle RSS (MiB) | 7.45 | Valid | — |
| OpenAI Responses → Anthropic | Steady-state RSS (MiB) | 20.87 | Valid | — |
| OpenAI Responses → Anthropic | Recovered RSS (MiB) | 15.52 | Valid | — |
| OpenAI Responses → Anthropic | Peak RSS (MiB) | 21.23 | Valid | — |
| OpenAI Responses → Anthropic | Peak RSS high-water mark (MiB) | 20.79 | Valid | — |
| OpenAI Responses → Anthropic | Time to plateau (s) | — | Not measured | — |
| OpenAI Responses → Anthropic | Memory growth (MiB/min) | -0.131 | Valid | — |
| OpenAI Responses → Anthropic | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → Anthropic | Memory shape under load | 0 | Valid | — |
| OpenAI Responses → Anthropic | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → Anthropic | Idle memory static | 0 | Valid | — |
| OpenAI Responses → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,647 | concurrency 16; n=177,901 | 1 ms |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,773 | concurrency 32; n=178,657 | 5 ms |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,773 | concurrency 32; n=178,657 | 10 ms |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,773 | concurrency 32; n=178,657 | 50 ms |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,773 | concurrency 32; n=178,657 | 100 ms |
| OpenAI Responses → Anthropic | Throughput (req/s) | 29,773 | concurrency 32; n=178,657 | Unbounded |
| Anthropic → Anthropic | Added latency p50 (µs) | 85 | Valid · n=55,235 gateway / 261,103 direct | — |
| Anthropic → Anthropic | Added latency p99 (µs) | 96 | Valid · n=55,235 gateway / 261,103 direct | — |
| Anthropic → Anthropic | Gateway p99 at concurrency 1 (µs) | 124 | Valid · n=55,235 | — |
| Anthropic → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=261,103 | — |
| Anthropic → Anthropic | CPU per request (µs) | 74.12 | Valid | — |
| Anthropic → Anthropic | Requests per CPU-second | 13,491.6 | Valid | — |
| Anthropic → Anthropic | Cost window concurrency | 8 | Valid | — |
| Anthropic → Anthropic | Cost window core utilisation | 0.957 | Valid | — |
| Anthropic → Anthropic | Cost window successful requests | 310,846 | Valid | — |
| Anthropic → Anthropic | Cost window throughput (req/s) | 51,801 | Valid | — |
| Anthropic → Anthropic | Cost window threads | 10 | Valid | — |
| Anthropic → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → Anthropic | Cost window major page faults | 0 | Valid | — |
| Anthropic → Anthropic | Added time to first event p50 (µs) | 125 | Valid · n=100 gateway / 100 direct | — |
| Anthropic → Anthropic | Added time to first event p99 (µs) | 229 | Valid · n=100 gateway / 100 direct | — |
| Anthropic → Anthropic | Added inter-frame gap p50 (µs) | 2 | Valid · n=63 gateway / 63 direct | — |
| Anthropic → Anthropic | Added inter-frame gap p99 (µs) | — | Below resolution · n=63 gateway / 63 direct | — |
| Anthropic → Anthropic | Streams sustained (frames/s) | 43,414.8 | Valid | — |
| Anthropic → Anthropic | Streams sustained | 1,698 | Valid | — |
| Anthropic → Anthropic | Idle RSS (MiB) | 7.44 | Valid | — |
| Anthropic → Anthropic | Steady-state RSS (MiB) | 22.21 | Valid | — |
| Anthropic → Anthropic | Recovered RSS (MiB) | 16.39 | Valid | — |
| Anthropic → Anthropic | Peak RSS (MiB) | 23.27 | Valid | — |
| Anthropic → Anthropic | Peak RSS high-water mark (MiB) | 22.98 | Valid | — |
| Anthropic → Anthropic | Time to plateau (s) | — | Not measured | — |
| Anthropic → Anthropic | Memory growth (MiB/min) | 0.584 | Valid | — |
| Anthropic → Anthropic | Memory load duration (s) | 300 | Valid | — |
| Anthropic → Anthropic | Memory shape under load | 1 | Valid | — |
| Anthropic → Anthropic | Idle memory shape | 0 | Valid | — |
| Anthropic → Anthropic | Idle memory static | 0 | Valid | — |
| Anthropic → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → Anthropic | Throughput (req/s) | 57,786 | concurrency 32; n=346,749 | 1 ms |
| Anthropic → Anthropic | Throughput (req/s) | 58,879 | concurrency 128; n=353,323 | 5 ms |
| Anthropic → Anthropic | Throughput (req/s) | 58,879 | concurrency 128; n=353,323 | 10 ms |
| Anthropic → Anthropic | Throughput (req/s) | 58,879 | concurrency 128; n=353,323 | 50 ms |
| Anthropic → Anthropic | Throughput (req/s) | 58,879 | concurrency 128; n=353,323 | 100 ms |
| Anthropic → Anthropic | Throughput (req/s) | 58,879 | concurrency 128; n=353,323 | Unbounded |
| Gemini → Anthropic | Added latency p50 (µs) | 105 | Valid · n=46,775 gateway / 263,231 direct | — |
| Gemini → Anthropic | Added latency p99 (µs) | 119 | Valid · n=46,775 gateway / 263,231 direct | — |
| Gemini → Anthropic | Gateway p99 at concurrency 1 (µs) | 146 | Valid · n=46,775 | — |
| Gemini → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=263,231 | — |
| Gemini → Anthropic | CPU per request (µs) | 93.11 | Valid | — |
| Gemini → Anthropic | Requests per CPU-second | 10,739.9 | Valid | — |
| Gemini → Anthropic | Cost window concurrency | 8 | Valid | — |
| Gemini → Anthropic | Cost window core utilisation | 0.967 | Valid | — |
| Gemini → Anthropic | Cost window successful requests | 250,132 | Valid | — |
| Gemini → Anthropic | Cost window throughput (req/s) | 41,684 | Valid | — |
| Gemini → Anthropic | Cost window threads | 10 | Valid | — |
| Gemini → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → Anthropic | Cost window major page faults | 0 | Valid | — |
| Gemini → Anthropic | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → Anthropic | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → Anthropic | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → Anthropic | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → Anthropic | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → Anthropic | Streams sustained | — | Untestable | — |
| Gemini → Anthropic | Idle RSS (MiB) | 7.43 | Valid | — |
| Gemini → Anthropic | Steady-state RSS (MiB) | 22.05 | Valid | — |
| Gemini → Anthropic | Recovered RSS (MiB) | 16.32 | Valid | — |
| Gemini → Anthropic | Peak RSS (MiB) | 22.44 | Valid | — |
| Gemini → Anthropic | Peak RSS high-water mark (MiB) | 22.28 | Valid | — |
| Gemini → Anthropic | Time to plateau (s) | — | Not measured | — |
| Gemini → Anthropic | Memory growth (MiB/min) | 0.159 | Valid | — |
| Gemini → Anthropic | Memory load duration (s) | 300 | Valid | — |
| Gemini → Anthropic | Memory shape under load | 0 | Valid | — |
| Gemini → Anthropic | Idle memory shape | 0 | Valid | — |
| Gemini → Anthropic | Idle memory static | 0 | Valid | — |
| Gemini → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → Anthropic | Throughput (req/s) | 44,397 | concurrency 16; n=266,417 | 1 ms |
| Gemini → Anthropic | Throughput (req/s) | 45,198 | concurrency 32; n=271,212 | 5 ms |
| Gemini → Anthropic | Throughput (req/s) | 45,198 | concurrency 32; n=271,212 | 10 ms |
| Gemini → Anthropic | Throughput (req/s) | 45,198 | concurrency 32; n=271,212 | 50 ms |
| Gemini → Anthropic | Throughput (req/s) | 45,198 | concurrency 32; n=271,212 | 100 ms |
| Gemini → Anthropic | Throughput (req/s) | 45,198 | concurrency 32; n=271,212 | Unbounded |
| Bedrock → Anthropic | Added latency p50 (µs) | 106 | Valid · n=46,122 gateway / 266,395 direct | — |
| Bedrock → Anthropic | Added latency p99 (µs) | 120 | Valid · n=46,122 gateway / 266,395 direct | — |
| Bedrock → Anthropic | Gateway p99 at concurrency 1 (µs) | 147 | Valid · n=46,122 | — |
| Bedrock → Anthropic | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=266,395 | — |
| Bedrock → Anthropic | CPU per request (µs) | 93.72 | Valid | — |
| Bedrock → Anthropic | Requests per CPU-second | 10,670.4 | Valid | — |
| Bedrock → Anthropic | Cost window concurrency | 8 | Valid | — |
| Bedrock → Anthropic | Cost window core utilisation | 0.965 | Valid | — |
| Bedrock → Anthropic | Cost window successful requests | 248,087 | Valid | — |
| Bedrock → Anthropic | Cost window throughput (req/s) | 41,343 | Valid | — |
| Bedrock → Anthropic | Cost window threads | 10 | Valid | — |
| Bedrock → Anthropic | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → Anthropic | Cost window major page faults | 0 | Valid | — |
| Bedrock → Anthropic | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → Anthropic | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → Anthropic | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → Anthropic | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → Anthropic | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → Anthropic | Streams sustained | — | Untestable | — |
| Bedrock → Anthropic | Idle RSS (MiB) | 7.44 | Valid | — |
| Bedrock → Anthropic | Steady-state RSS (MiB) | 26.27 | Valid | — |
| Bedrock → Anthropic | Recovered RSS (MiB) | 16.05 | Valid | — |
| Bedrock → Anthropic | Peak RSS (MiB) | 26.76 | Valid | — |
| Bedrock → Anthropic | Peak RSS high-water mark (MiB) | 26.53 | Valid | — |
| Bedrock → Anthropic | Time to plateau (s) | — | Not measured | — |
| Bedrock → Anthropic | Memory growth (MiB/min) | 0.442 | Valid | — |
| Bedrock → Anthropic | Memory load duration (s) | 300 | Valid | — |
| Bedrock → Anthropic | Memory shape under load | 0 | Valid | — |
| Bedrock → Anthropic | Idle memory shape | 0 | Valid | — |
| Bedrock → Anthropic | Idle memory static | 0 | Valid | — |
| Bedrock → Anthropic | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → Anthropic | Throughput (req/s) | 43,964 | concurrency 16; n=263,811 | 1 ms |
| Bedrock → Anthropic | Throughput (req/s) | 45,086 | concurrency 64; n=270,545 | 5 ms |
| Bedrock → Anthropic | Throughput (req/s) | 45,086 | concurrency 64; n=270,545 | 10 ms |
| Bedrock → Anthropic | Throughput (req/s) | 45,086 | concurrency 64; n=270,545 | 50 ms |
| Bedrock → Anthropic | Throughput (req/s) | 45,086 | concurrency 64; n=270,545 | 100 ms |
| Bedrock → Anthropic | Throughput (req/s) | 45,086 | concurrency 64; n=270,545 | Unbounded |
| OpenAI Responses → Bedrock | Added latency p50 (µs) | 171 | Valid · n=30,814 gateway / 264,867 direct | — |
| OpenAI Responses → Bedrock | Added latency p99 (µs) | 192 | Valid · n=30,814 gateway / 264,867 direct | — |
| OpenAI Responses → Bedrock | Gateway p99 at concurrency 1 (µs) | 219 | Valid · n=30,814 | — |
| OpenAI Responses → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=264,867 | — |
| OpenAI Responses → Bedrock | CPU per request (µs) | 157.1 | Valid | — |
| OpenAI Responses → Bedrock | Requests per CPU-second | 6,363.6 | Valid | — |
| OpenAI Responses → Bedrock | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → Bedrock | Cost window core utilisation | 0.979 | Valid | — |
| OpenAI Responses → Bedrock | Cost window successful requests | 149,989 | Valid | — |
| OpenAI Responses → Bedrock | Cost window throughput (req/s) | 24,995 | Valid | — |
| OpenAI Responses → Bedrock | Cost window threads | 10 | Valid | — |
| OpenAI Responses → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → Bedrock | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → Bedrock | Streams sustained | — | Untestable | — |
| OpenAI Responses → Bedrock | Idle RSS (MiB) | 7.44 | Valid | — |
| OpenAI Responses → Bedrock | Steady-state RSS (MiB) | 23.22 | Valid | — |
| OpenAI Responses → Bedrock | Recovered RSS (MiB) | 14.27 | Valid | — |
| OpenAI Responses → Bedrock | Peak RSS (MiB) | 23.45 | Valid | — |
| OpenAI Responses → Bedrock | Peak RSS high-water mark (MiB) | 23.23 | Valid | — |
| OpenAI Responses → Bedrock | Time to plateau (s) | 300 | Valid | — |
| OpenAI Responses → Bedrock | Memory growth (MiB/min) | 0.222 | Valid | — |
| OpenAI Responses → Bedrock | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → Bedrock | Memory shape under load | — | Not measured | — |
| OpenAI Responses → Bedrock | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → Bedrock | Idle memory static | 0 | Valid | — |
| OpenAI Responses → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | 1 ms |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | 5 ms |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | 10 ms |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | 50 ms |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | 100 ms |
| OpenAI Responses → Bedrock | Throughput (req/s) | 25,676 | concurrency 16; n=154,074 | Unbounded |
| Gemini → Bedrock | Added latency p50 (µs) | 129 | Valid · n=39,387 gateway / 270,577 direct | — |
| Gemini → Bedrock | Added latency p99 (µs) | 148 | Valid · n=39,387 gateway / 270,577 direct | — |
| Gemini → Bedrock | Gateway p99 at concurrency 1 (µs) | 175 | Valid · n=39,387 | — |
| Gemini → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=270,577 | — |
| Gemini → Bedrock | CPU per request (µs) | 115.7 | Valid | — |
| Gemini → Bedrock | Requests per CPU-second | 8,644.8 | Valid | — |
| Gemini → Bedrock | Cost window concurrency | 8 | Valid | — |
| Gemini → Bedrock | Cost window core utilisation | 0.969 | Valid | — |
| Gemini → Bedrock | Cost window successful requests | 201,769 | Valid | — |
| Gemini → Bedrock | Cost window throughput (req/s) | 33,624 | Valid | — |
| Gemini → Bedrock | Cost window threads | 10 | Valid | — |
| Gemini → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → Bedrock | Cost window major page faults | 0 | Valid | — |
| Gemini → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → Bedrock | Streams sustained | — | Untestable | — |
| Gemini → Bedrock | Idle RSS (MiB) | 7.45 | Valid | — |
| Gemini → Bedrock | Steady-state RSS (MiB) | 21.81 | Valid | — |
| Gemini → Bedrock | Recovered RSS (MiB) | 15.04 | Valid | — |
| Gemini → Bedrock | Peak RSS (MiB) | 22.15 | Valid | — |
| Gemini → Bedrock | Peak RSS high-water mark (MiB) | 21.81 | Valid | — |
| Gemini → Bedrock | Time to plateau (s) | — | Not measured | — |
| Gemini → Bedrock | Memory growth (MiB/min) | 0.346 | Valid | — |
| Gemini → Bedrock | Memory load duration (s) | 300 | Valid | — |
| Gemini → Bedrock | Memory shape under load | 0 | Valid | — |
| Gemini → Bedrock | Idle memory shape | 0 | Valid | — |
| Gemini → Bedrock | Idle memory static | 0 | Valid | — |
| Gemini → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → Bedrock | Throughput (req/s) | 34,935 | concurrency 16; n=209,641 | 1 ms |
| Gemini → Bedrock | Throughput (req/s) | 35,197 | concurrency 64; n=211,190 | 5 ms |
| Gemini → Bedrock | Throughput (req/s) | 35,197 | concurrency 64; n=211,190 | 10 ms |
| Gemini → Bedrock | Throughput (req/s) | 35,197 | concurrency 64; n=211,190 | 50 ms |
| Gemini → Bedrock | Throughput (req/s) | 35,197 | concurrency 64; n=211,190 | 100 ms |
| Gemini → Bedrock | Throughput (req/s) | 35,197 | concurrency 64; n=211,190 | Unbounded |
| Cohere → Bedrock | Added latency p50 (µs) | 120 | Valid · n=41,840 gateway / 268,504 direct | — |
| Cohere → Bedrock | Added latency p99 (µs) | 136 | Valid · n=41,840 gateway / 268,504 direct | — |
| Cohere → Bedrock | Gateway p99 at concurrency 1 (µs) | 163 | Valid · n=41,840 | — |
| Cohere → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=268,504 | — |
| Cohere → Bedrock | CPU per request (µs) | 109.9 | Valid | — |
| Cohere → Bedrock | Requests per CPU-second | 9,099.7 | Valid | — |
| Cohere → Bedrock | Cost window concurrency | 8 | Valid | — |
| Cohere → Bedrock | Cost window core utilisation | 0.968 | Valid | — |
| Cohere → Bedrock | Cost window successful requests | 212,113 | Valid | — |
| Cohere → Bedrock | Cost window throughput (req/s) | 35,348 | Valid | — |
| Cohere → Bedrock | Cost window threads | 10 | Valid | — |
| Cohere → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → Bedrock | Cost window major page faults | 0 | Valid | — |
| Cohere → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → Bedrock | Streams sustained | — | Untestable | — |
| Cohere → Bedrock | Idle RSS (MiB) | 7.42 | Valid | — |
| Cohere → Bedrock | Steady-state RSS (MiB) | 20.3 | Valid | — |
| Cohere → Bedrock | Recovered RSS (MiB) | 15.18 | Valid | — |
| Cohere → Bedrock | Peak RSS (MiB) | 20.46 | Valid | — |
| Cohere → Bedrock | Peak RSS high-water mark (MiB) | 20.18 | Valid | — |
| Cohere → Bedrock | Time to plateau (s) | — | Not measured | — |
| Cohere → Bedrock | Memory growth (MiB/min) | 0.344 | Valid | — |
| Cohere → Bedrock | Memory load duration (s) | 300 | Valid | — |
| Cohere → Bedrock | Memory shape under load | 0 | Valid | — |
| Cohere → Bedrock | Idle memory shape | 0 | Valid | — |
| Cohere → Bedrock | Idle memory static | 0 | Valid | — |
| Cohere → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → Bedrock | Throughput (req/s) | 36,699 | concurrency 16; n=220,217 | 1 ms |
| Cohere → Bedrock | Throughput (req/s) | 36,731 | concurrency 32; n=220,409 | 5 ms |
| Cohere → Bedrock | Throughput (req/s) | 36,731 | concurrency 32; n=220,409 | 10 ms |
| Cohere → Bedrock | Throughput (req/s) | 36,731 | concurrency 32; n=220,409 | 50 ms |
| Cohere → Bedrock | Throughput (req/s) | 36,731 | concurrency 32; n=220,409 | 100 ms |
| Cohere → Bedrock | Throughput (req/s) | 36,731 | concurrency 32; n=220,409 | Unbounded |
| Bedrock → Bedrock | Added latency p50 (µs) | 110 | Valid · n=44,888 gateway / 266,888 direct | — |
| Bedrock → Bedrock | Added latency p99 (µs) | 127 | Valid · n=44,888 gateway / 266,888 direct | — |
| Bedrock → Bedrock | Gateway p99 at concurrency 1 (µs) | 154 | Valid · n=44,888 | — |
| Bedrock → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=266,888 | — |
| Bedrock → Bedrock | CPU per request (µs) | 101.1 | Valid | — |
| Bedrock → Bedrock | Requests per CPU-second | 9,893.4 | Valid | — |
| Bedrock → Bedrock | Cost window concurrency | 8 | Valid | — |
| Bedrock → Bedrock | Cost window core utilisation | 0.966 | Valid | — |
| Bedrock → Bedrock | Cost window successful requests | 230,120 | Valid | — |
| Bedrock → Bedrock | Cost window throughput (req/s) | 38,348 | Valid | — |
| Bedrock → Bedrock | Cost window threads | 10 | Valid | — |
| Bedrock → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → Bedrock | Cost window major page faults | 0 | Valid | — |
| Bedrock → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → Bedrock | Streams sustained | — | Untestable | — |
| Bedrock → Bedrock | Idle RSS (MiB) | 7.42 | Valid | — |
| Bedrock → Bedrock | Steady-state RSS (MiB) | 20.46 | Valid | — |
| Bedrock → Bedrock | Recovered RSS (MiB) | 15.18 | Valid | — |
| Bedrock → Bedrock | Peak RSS (MiB) | 20.75 | Valid | — |
| Bedrock → Bedrock | Peak RSS high-water mark (MiB) | 20.65 | Valid | — |
| Bedrock → Bedrock | Time to plateau (s) | — | Not measured | — |
| Bedrock → Bedrock | Memory growth (MiB/min) | -0.062 | Valid | — |
| Bedrock → Bedrock | Memory load duration (s) | 300 | Valid | — |
| Bedrock → Bedrock | Memory shape under load | 0 | Valid | — |
| Bedrock → Bedrock | Idle memory shape | 0 | Valid | — |
| Bedrock → Bedrock | Idle memory static | 0 | Valid | — |
| Bedrock → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → Bedrock | Throughput (req/s) | 39,956 | concurrency 16; n=239,762 | 1 ms |
| Bedrock → Bedrock | Throughput (req/s) | 40,107 | concurrency 32; n=240,666 | 5 ms |
| Bedrock → Bedrock | Throughput (req/s) | 40,107 | concurrency 32; n=240,666 | 10 ms |
| Bedrock → Bedrock | Throughput (req/s) | 40,107 | concurrency 32; n=240,666 | 50 ms |
| Bedrock → Bedrock | Throughput (req/s) | 40,107 | concurrency 32; n=240,666 | 100 ms |
| Bedrock → Bedrock | Throughput (req/s) | 40,107 | concurrency 32; n=240,666 | Unbounded |
| OpenAI → Bedrock | Added latency p50 (µs) | 125 | Valid · n=40,371 gateway / 265,424 direct | — |
| OpenAI → Bedrock | Added latency p99 (µs) | 142 | Valid · n=40,371 gateway / 265,424 direct | — |
| OpenAI → Bedrock | Gateway p99 at concurrency 1 (µs) | 170 | Valid · n=40,371 | — |
| OpenAI → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=265,424 | — |
| OpenAI → Bedrock | CPU per request (µs) | 111.4 | Valid | — |
| OpenAI → Bedrock | Requests per CPU-second | 8,978.8 | Valid | — |
| OpenAI → Bedrock | Cost window concurrency | 8 | Valid | — |
| OpenAI → Bedrock | Cost window core utilisation | 0.971 | Valid | — |
| OpenAI → Bedrock | Cost window successful requests | 209,835 | Valid | — |
| OpenAI → Bedrock | Cost window throughput (req/s) | 34,969 | Valid | — |
| OpenAI → Bedrock | Cost window threads | 10 | Valid | — |
| OpenAI → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → Bedrock | Cost window major page faults | 0 | Valid | — |
| OpenAI → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI → Bedrock | Streams sustained | — | Untestable | — |
| OpenAI → Bedrock | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI → Bedrock | Steady-state RSS (MiB) | 19.93 | Valid | — |
| OpenAI → Bedrock | Recovered RSS (MiB) | 15.18 | Valid | — |
| OpenAI → Bedrock | Peak RSS (MiB) | 20.37 | Valid | — |
| OpenAI → Bedrock | Peak RSS high-water mark (MiB) | 19.97 | Valid | — |
| OpenAI → Bedrock | Time to plateau (s) | — | Not measured | — |
| OpenAI → Bedrock | Memory growth (MiB/min) | -0.073 | Valid | — |
| OpenAI → Bedrock | Memory load duration (s) | 300 | Valid | — |
| OpenAI → Bedrock | Memory shape under load | 0 | Valid | — |
| OpenAI → Bedrock | Idle memory shape | 0 | Valid | — |
| OpenAI → Bedrock | Idle memory static | 0 | Valid | — |
| OpenAI → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → Bedrock | Throughput (req/s) | 36,468 | concurrency 16; n=218,834 | 1 ms |
| OpenAI → Bedrock | Throughput (req/s) | 36,523 | concurrency 32; n=219,157 | 5 ms |
| OpenAI → Bedrock | Throughput (req/s) | 36,523 | concurrency 32; n=219,157 | 10 ms |
| OpenAI → Bedrock | Throughput (req/s) | 36,523 | concurrency 32; n=219,157 | 50 ms |
| OpenAI → Bedrock | Throughput (req/s) | 36,523 | concurrency 32; n=219,157 | 100 ms |
| OpenAI → Bedrock | Throughput (req/s) | 36,523 | concurrency 32; n=219,157 | Unbounded |
| Anthropic → Bedrock | Added latency p50 (µs) | 123 | Valid · n=40,751 gateway / 263,427 direct | — |
| Anthropic → Bedrock | Added latency p99 (µs) | 140 | Valid · n=40,751 gateway / 263,427 direct | — |
| Anthropic → Bedrock | Gateway p99 at concurrency 1 (µs) | 168 | Valid · n=40,751 | — |
| Anthropic → Bedrock | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=263,427 | — |
| Anthropic → Bedrock | CPU per request (µs) | 111.5 | Valid | — |
| Anthropic → Bedrock | Requests per CPU-second | 8,967.3 | Valid | — |
| Anthropic → Bedrock | Cost window concurrency | 8 | Valid | — |
| Anthropic → Bedrock | Cost window core utilisation | 0.969 | Valid | — |
| Anthropic → Bedrock | Cost window successful requests | 209,296 | Valid | — |
| Anthropic → Bedrock | Cost window throughput (req/s) | 34,878 | Valid | — |
| Anthropic → Bedrock | Cost window threads | 10 | Valid | — |
| Anthropic → Bedrock | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → Bedrock | Cost window major page faults | 0 | Valid | — |
| Anthropic → Bedrock | Added time to first event p50 (µs) | — | Untestable | — |
| Anthropic → Bedrock | Added time to first event p99 (µs) | — | Untestable | — |
| Anthropic → Bedrock | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Anthropic → Bedrock | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Anthropic → Bedrock | Streams sustained (frames/s) | — | Untestable | — |
| Anthropic → Bedrock | Streams sustained | — | Untestable | — |
| Anthropic → Bedrock | Idle RSS (MiB) | 7.42 | Valid | — |
| Anthropic → Bedrock | Steady-state RSS (MiB) | 20.27 | Valid | — |
| Anthropic → Bedrock | Recovered RSS (MiB) | 15.97 | Valid | — |
| Anthropic → Bedrock | Peak RSS (MiB) | 21.07 | Valid | — |
| Anthropic → Bedrock | Peak RSS high-water mark (MiB) | 21.01 | Valid | — |
| Anthropic → Bedrock | Time to plateau (s) | — | Not measured | — |
| Anthropic → Bedrock | Memory growth (MiB/min) | 0.559 | Valid | — |
| Anthropic → Bedrock | Memory load duration (s) | 300 | Valid | — |
| Anthropic → Bedrock | Memory shape under load | 1 | Valid | — |
| Anthropic → Bedrock | Idle memory shape | 0 | Valid | — |
| Anthropic → Bedrock | Idle memory static | 0 | Valid | — |
| Anthropic → Bedrock | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → Bedrock | Throughput (req/s) | 36,181 | concurrency 16; n=217,103 | 1 ms |
| Anthropic → Bedrock | Throughput (req/s) | 36,453 | concurrency 64; n=218,731 | 5 ms |
| Anthropic → Bedrock | Throughput (req/s) | 36,453 | concurrency 64; n=218,731 | 10 ms |
| Anthropic → Bedrock | Throughput (req/s) | 36,453 | concurrency 64; n=218,731 | 50 ms |
| Anthropic → Bedrock | Throughput (req/s) | 36,453 | concurrency 64; n=218,731 | 100 ms |
| Anthropic → Bedrock | Throughput (req/s) | 36,453 | concurrency 64; n=218,731 | Unbounded |
| OpenAI → OpenAI | Added latency p50 (µs) | 78 | Valid · n=59,266 gateway / 266,330 direct | — |
| OpenAI → OpenAI | Added latency p99 (µs) | 89 | Valid · n=59,266 gateway / 266,330 direct | — |
| OpenAI → OpenAI | Gateway p99 at concurrency 1 (µs) | 117 | Valid · n=59,266 | — |
| OpenAI → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=266,330 | — |
| OpenAI → OpenAI | CPU per request (µs) | 68.74 | Valid | — |
| OpenAI → OpenAI | Requests per CPU-second | 14,547.2 | Valid | — |
| OpenAI → OpenAI | Cost window concurrency | 8 | Valid | — |
| OpenAI → OpenAI | Cost window core utilisation | 0.957 | Valid | — |
| OpenAI → OpenAI | Cost window successful requests | 335,167 | Valid | — |
| OpenAI → OpenAI | Cost window throughput (req/s) | 55,854 | Valid | — |
| OpenAI → OpenAI | Cost window threads | 10 | Valid | — |
| OpenAI → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → OpenAI | Cost window major page faults | 0 | Valid | — |
| OpenAI → OpenAI | Added time to first event p50 (µs) | 118 | Valid · n=100 gateway / 100 direct | — |
| OpenAI → OpenAI | Added time to first event p99 (µs) | 254 | Valid · n=100 gateway / 100 direct | — |
| OpenAI → OpenAI | Added inter-frame gap p50 (µs) | — | Below resolution · n=63 gateway / 63 direct | — |
| OpenAI → OpenAI | Added inter-frame gap p99 (µs) | — | Below resolution · n=63 gateway / 63 direct | — |
| OpenAI → OpenAI | Streams sustained (frames/s) | 39,240.2 | Valid | — |
| OpenAI → OpenAI | Streams sustained | 1,452 | Valid | — |
| OpenAI → OpenAI | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI → OpenAI | Steady-state RSS (MiB) | 21.49 | Valid | — |
| OpenAI → OpenAI | Recovered RSS (MiB) | 15.7 | Valid | — |
| OpenAI → OpenAI | Peak RSS (MiB) | 22.35 | Valid | — |
| OpenAI → OpenAI | Peak RSS high-water mark (MiB) | 22.27 | Valid | — |
| OpenAI → OpenAI | Time to plateau (s) | — | Not measured | — |
| OpenAI → OpenAI | Memory growth (MiB/min) | 0.391 | Valid | — |
| OpenAI → OpenAI | Memory load duration (s) | 300 | Valid | — |
| OpenAI → OpenAI | Memory shape under load | 1 | Valid | — |
| OpenAI → OpenAI | Idle memory shape | 0 | Valid | — |
| OpenAI → OpenAI | Idle memory static | 0 | Valid | — |
| OpenAI → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → OpenAI | Throughput (req/s) | 62,231 | concurrency 32; n=373,417 | 1 ms |
| OpenAI → OpenAI | Throughput (req/s) | 62,587 | concurrency 128; n=375,578 | 5 ms |
| OpenAI → OpenAI | Throughput (req/s) | 62,587 | concurrency 128; n=375,578 | 10 ms |
| OpenAI → OpenAI | Throughput (req/s) | 62,587 | concurrency 128; n=375,578 | 50 ms |
| OpenAI → OpenAI | Throughput (req/s) | 62,587 | concurrency 128; n=375,578 | 100 ms |
| OpenAI → OpenAI | Throughput (req/s) | 62,587 | concurrency 128; n=375,578 | Unbounded |
| Cohere → OpenAI | Added latency p50 (µs) | 98 | Valid · n=49,364 gateway / 265,547 direct | — |
| Cohere → OpenAI | Added latency p99 (µs) | 111 | Valid · n=49,364 gateway / 265,547 direct | — |
| Cohere → OpenAI | Gateway p99 at concurrency 1 (µs) | 138 | Valid · n=49,364 | — |
| Cohere → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=265,547 | — |
| Cohere → OpenAI | CPU per request (µs) | 86.9 | Valid | — |
| Cohere → OpenAI | Requests per CPU-second | 11,507.9 | Valid | — |
| Cohere → OpenAI | Cost window concurrency | 8 | Valid | — |
| Cohere → OpenAI | Cost window core utilisation | 0.963 | Valid | — |
| Cohere → OpenAI | Cost window successful requests | 266,754 | Valid | — |
| Cohere → OpenAI | Cost window throughput (req/s) | 44,454 | Valid | — |
| Cohere → OpenAI | Cost window threads | 10 | Valid | — |
| Cohere → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → OpenAI | Cost window major page faults | 0 | Valid | — |
| Cohere → OpenAI | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → OpenAI | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → OpenAI | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → OpenAI | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → OpenAI | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → OpenAI | Streams sustained | — | Untestable | — |
| Cohere → OpenAI | Idle RSS (MiB) | 7.43 | Valid | — |
| Cohere → OpenAI | Steady-state RSS (MiB) | 20.19 | Valid | — |
| Cohere → OpenAI | Recovered RSS (MiB) | 15.3 | Valid | — |
| Cohere → OpenAI | Peak RSS (MiB) | 20.41 | Valid | — |
| Cohere → OpenAI | Peak RSS high-water mark (MiB) | 20.15 | Valid | — |
| Cohere → OpenAI | Time to plateau (s) | 276 | Valid | — |
| Cohere → OpenAI | Memory growth (MiB/min) | -0.038 | Valid | — |
| Cohere → OpenAI | Memory load duration (s) | 300 | Valid | — |
| Cohere → OpenAI | Memory shape under load | — | Not measured | — |
| Cohere → OpenAI | Idle memory shape | 0 | Valid | — |
| Cohere → OpenAI | Idle memory static | 0 | Valid | — |
| Cohere → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → OpenAI | Throughput (req/s) | 47,293 | concurrency 16; n=283,788 | 1 ms |
| Cohere → OpenAI | Throughput (req/s) | 48,545 | concurrency 128; n=291,307 | 5 ms |
| Cohere → OpenAI | Throughput (req/s) | 48,545 | concurrency 128; n=291,307 | 10 ms |
| Cohere → OpenAI | Throughput (req/s) | 48,545 | concurrency 128; n=291,307 | 50 ms |
| Cohere → OpenAI | Throughput (req/s) | 48,545 | concurrency 128; n=291,307 | 100 ms |
| Cohere → OpenAI | Throughput (req/s) | 48,545 | concurrency 128; n=291,307 | Unbounded |
| Anthropic → OpenAI | Added latency p50 (µs) | 100 | Valid · n=48,328 gateway / 264,163 direct | — |
| Anthropic → OpenAI | Added latency p99 (µs) | 113 | Valid · n=48,328 gateway / 264,163 direct | — |
| Anthropic → OpenAI | Gateway p99 at concurrency 1 (µs) | 141 | Valid · n=48,328 | — |
| Anthropic → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=264,163 | — |
| Anthropic → OpenAI | CPU per request (µs) | 88.98 | Valid | — |
| Anthropic → OpenAI | Requests per CPU-second | 11,237.9 | Valid | — |
| Anthropic → OpenAI | Cost window concurrency | 8 | Valid | — |
| Anthropic → OpenAI | Cost window core utilisation | 0.963 | Valid | — |
| Anthropic → OpenAI | Cost window successful requests | 260,495 | Valid | — |
| Anthropic → OpenAI | Cost window throughput (req/s) | 43,410 | Valid | — |
| Anthropic → OpenAI | Cost window threads | 10 | Valid | — |
| Anthropic → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → OpenAI | Cost window major page faults | 0 | Valid | — |
| Anthropic → OpenAI | Added time to first event p50 (µs) | 118 | Valid · n=100 gateway / 100 direct | — |
| Anthropic → OpenAI | Added time to first event p99 (µs) | 213 | Valid · n=100 gateway / 100 direct | — |
| Anthropic → OpenAI | Added inter-frame gap p50 (µs) | 0 | Valid · n=63 gateway / 63 direct | — |
| Anthropic → OpenAI | Added inter-frame gap p99 (µs) | — | Below resolution · n=63 gateway / 63 direct | — |
| Anthropic → OpenAI | Streams sustained (frames/s) | 44,699.9 | Valid | — |
| Anthropic → OpenAI | Streams sustained | 1,610 | Valid | — |
| Anthropic → OpenAI | Idle RSS (MiB) | 7.45 | Valid | — |
| Anthropic → OpenAI | Steady-state RSS (MiB) | 23.48 | Valid | — |
| Anthropic → OpenAI | Recovered RSS (MiB) | 16.34 | Valid | — |
| Anthropic → OpenAI | Peak RSS (MiB) | 23.84 | Valid | — |
| Anthropic → OpenAI | Peak RSS high-water mark (MiB) | 23.72 | Valid | — |
| Anthropic → OpenAI | Time to plateau (s) | — | Not measured | — |
| Anthropic → OpenAI | Memory growth (MiB/min) | 0.3 | Valid | — |
| Anthropic → OpenAI | Memory load duration (s) | 300 | Valid | — |
| Anthropic → OpenAI | Memory shape under load | 0 | Valid | — |
| Anthropic → OpenAI | Idle memory shape | 0 | Valid | — |
| Anthropic → OpenAI | Idle memory static | 0 | Valid | — |
| Anthropic → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → OpenAI | Throughput (req/s) | 46,675 | concurrency 16; n=280,076 | 1 ms |
| Anthropic → OpenAI | Throughput (req/s) | 47,965 | concurrency 64; n=287,798 | 5 ms |
| Anthropic → OpenAI | Throughput (req/s) | 47,965 | concurrency 64; n=287,798 | 10 ms |
| Anthropic → OpenAI | Throughput (req/s) | 47,965 | concurrency 64; n=287,798 | 50 ms |
| Anthropic → OpenAI | Throughput (req/s) | 47,965 | concurrency 64; n=287,798 | 100 ms |
| Anthropic → OpenAI | Throughput (req/s) | 47,965 | concurrency 64; n=287,798 | Unbounded |
| Gemini → OpenAI | Added latency p50 (µs) | 104 | Valid · n=46,858 gateway / 265,285 direct | — |
| Gemini → OpenAI | Added latency p99 (µs) | 118 | Valid · n=46,858 gateway / 265,285 direct | — |
| Gemini → OpenAI | Gateway p99 at concurrency 1 (µs) | 145 | Valid · n=46,858 | — |
| Gemini → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=265,285 | — |
| Gemini → OpenAI | CPU per request (µs) | 92.72 | Valid | — |
| Gemini → OpenAI | Requests per CPU-second | 10,784.7 | Valid | — |
| Gemini → OpenAI | Cost window concurrency | 8 | Valid | — |
| Gemini → OpenAI | Cost window core utilisation | 0.961 | Valid | — |
| Gemini → OpenAI | Cost window successful requests | 249,665 | Valid | — |
| Gemini → OpenAI | Cost window throughput (req/s) | 41,606 | Valid | — |
| Gemini → OpenAI | Cost window threads | 10 | Valid | — |
| Gemini → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → OpenAI | Cost window major page faults | 0 | Valid | — |
| Gemini → OpenAI | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → OpenAI | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → OpenAI | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → OpenAI | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → OpenAI | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → OpenAI | Streams sustained | — | Untestable | — |
| Gemini → OpenAI | Idle RSS (MiB) | 7.43 | Valid | — |
| Gemini → OpenAI | Steady-state RSS (MiB) | 21.84 | Valid | — |
| Gemini → OpenAI | Recovered RSS (MiB) | 16.45 | Valid | — |
| Gemini → OpenAI | Peak RSS (MiB) | 22.51 | Valid | — |
| Gemini → OpenAI | Peak RSS high-water mark (MiB) | 22.48 | Valid | — |
| Gemini → OpenAI | Time to plateau (s) | — | Not measured | — |
| Gemini → OpenAI | Memory growth (MiB/min) | 0.572 | Valid | — |
| Gemini → OpenAI | Memory load duration (s) | 300 | Valid | — |
| Gemini → OpenAI | Memory shape under load | 1 | Valid | — |
| Gemini → OpenAI | Idle memory shape | 0 | Valid | — |
| Gemini → OpenAI | Idle memory static | 0 | Valid | — |
| Gemini → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → OpenAI | Throughput (req/s) | 45,104 | concurrency 16; n=270,658 | 1 ms |
| Gemini → OpenAI | Throughput (req/s) | 46,080 | concurrency 32; n=276,504 | 5 ms |
| Gemini → OpenAI | Throughput (req/s) | 46,080 | concurrency 32; n=276,504 | 10 ms |
| Gemini → OpenAI | Throughput (req/s) | 46,080 | concurrency 32; n=276,504 | 50 ms |
| Gemini → OpenAI | Throughput (req/s) | 46,080 | concurrency 32; n=276,504 | 100 ms |
| Gemini → OpenAI | Throughput (req/s) | 46,080 | concurrency 32; n=276,504 | Unbounded |
| OpenAI Responses → OpenAI | Added latency p50 (µs) | 144 | Valid · n=35,866 gateway / 266,530 direct | — |
| OpenAI Responses → OpenAI | Added latency p99 (µs) | 159 | Valid · n=35,866 gateway / 266,530 direct | — |
| OpenAI Responses → OpenAI | Gateway p99 at concurrency 1 (µs) | 186 | Valid · n=35,866 | — |
| OpenAI Responses → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=266,530 | — |
| OpenAI Responses → OpenAI | CPU per request (µs) | 134.5 | Valid | — |
| OpenAI Responses → OpenAI | Requests per CPU-second | 7,437 | Valid | — |
| OpenAI Responses → OpenAI | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → OpenAI | Cost window core utilisation | 0.981 | Valid | — |
| OpenAI Responses → OpenAI | Cost window successful requests | 175,663 | Valid | — |
| OpenAI Responses → OpenAI | Cost window throughput (req/s) | 29,273 | Valid | — |
| OpenAI Responses → OpenAI | Cost window threads | 10 | Valid | — |
| OpenAI Responses → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → OpenAI | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → OpenAI | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → OpenAI | Streams sustained | — | Untestable | — |
| OpenAI Responses → OpenAI | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI Responses → OpenAI | Steady-state RSS (MiB) | 21.13 | Valid | — |
| OpenAI Responses → OpenAI | Recovered RSS (MiB) | 15.38 | Valid | — |
| OpenAI Responses → OpenAI | Peak RSS (MiB) | 21.75 | Valid | — |
| OpenAI Responses → OpenAI | Peak RSS high-water mark (MiB) | 21.45 | Valid | — |
| OpenAI Responses → OpenAI | Time to plateau (s) | — | Not measured | — |
| OpenAI Responses → OpenAI | Memory growth (MiB/min) | -0.37 | Valid | — |
| OpenAI Responses → OpenAI | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → OpenAI | Memory shape under load | 0 | Valid | — |
| OpenAI Responses → OpenAI | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → OpenAI | Idle memory static | 0 | Valid | — |
| OpenAI Responses → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,322 | concurrency 16; n=181,950 | 1 ms |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,374 | concurrency 32; n=182,256 | 5 ms |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,374 | concurrency 32; n=182,256 | 10 ms |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,374 | concurrency 32; n=182,256 | 50 ms |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,374 | concurrency 32; n=182,256 | 100 ms |
| OpenAI Responses → OpenAI | Throughput (req/s) | 30,374 | concurrency 32; n=182,256 | Unbounded |
| Bedrock → OpenAI | Added latency p50 (µs) | 103 | Valid · n=47,501 gateway / 263,962 direct | — |
| Bedrock → OpenAI | Added latency p99 (µs) | 115 | Valid · n=47,501 gateway / 263,962 direct | — |
| Bedrock → OpenAI | Gateway p99 at concurrency 1 (µs) | 143 | Valid · n=47,501 | — |
| Bedrock → OpenAI | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=263,962 | — |
| Bedrock → OpenAI | CPU per request (µs) | 92.12 | Valid | — |
| Bedrock → OpenAI | Requests per CPU-second | 10,855.2 | Valid | — |
| Bedrock → OpenAI | Cost window concurrency | 8 | Valid | — |
| Bedrock → OpenAI | Cost window core utilisation | 0.961 | Valid | — |
| Bedrock → OpenAI | Cost window successful requests | 251,298 | Valid | — |
| Bedrock → OpenAI | Cost window throughput (req/s) | 41,879 | Valid | — |
| Bedrock → OpenAI | Cost window threads | 10 | Valid | — |
| Bedrock → OpenAI | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → OpenAI | Cost window major page faults | 0 | Valid | — |
| Bedrock → OpenAI | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → OpenAI | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → OpenAI | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → OpenAI | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → OpenAI | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → OpenAI | Streams sustained | — | Untestable | — |
| Bedrock → OpenAI | Idle RSS (MiB) | 7.42 | Valid | — |
| Bedrock → OpenAI | Steady-state RSS (MiB) | 20.69 | Valid | — |
| Bedrock → OpenAI | Recovered RSS (MiB) | 15.17 | Valid | — |
| Bedrock → OpenAI | Peak RSS (MiB) | 21.05 | Valid | — |
| Bedrock → OpenAI | Peak RSS high-water mark (MiB) | 20.8 | Valid | — |
| Bedrock → OpenAI | Time to plateau (s) | 192 | Valid | — |
| Bedrock → OpenAI | Memory growth (MiB/min) | -0.05 | Valid | — |
| Bedrock → OpenAI | Memory load duration (s) | 300 | Valid | — |
| Bedrock → OpenAI | Memory shape under load | — | Not measured | — |
| Bedrock → OpenAI | Idle memory shape | 0 | Valid | — |
| Bedrock → OpenAI | Idle memory static | 0 | Valid | — |
| Bedrock → OpenAI | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → OpenAI | Throughput (req/s) | 45,153 | concurrency 16; n=270,945 | 1 ms |
| Bedrock → OpenAI | Throughput (req/s) | 46,330 | concurrency 64; n=277,993 | 5 ms |
| Bedrock → OpenAI | Throughput (req/s) | 46,330 | concurrency 64; n=277,993 | 10 ms |
| Bedrock → OpenAI | Throughput (req/s) | 46,330 | concurrency 64; n=277,993 | 50 ms |
| Bedrock → OpenAI | Throughput (req/s) | 46,330 | concurrency 64; n=277,993 | 100 ms |
| Bedrock → OpenAI | Throughput (req/s) | 46,330 | concurrency 64; n=277,993 | Unbounded |
| Anthropic → Cohere | Added latency p50 (µs) | 99 | Valid · n=48,830 gateway / 265,034 direct | — |
| Anthropic → Cohere | Added latency p99 (µs) | 112 | Valid · n=48,830 gateway / 265,034 direct | — |
| Anthropic → Cohere | Gateway p99 at concurrency 1 (µs) | 139 | Valid · n=48,830 | — |
| Anthropic → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=265,034 | — |
| Anthropic → Cohere | CPU per request (µs) | 87.39 | Valid | — |
| Anthropic → Cohere | Requests per CPU-second | 11,443 | Valid | — |
| Anthropic → Cohere | Cost window concurrency | 8 | Valid | — |
| Anthropic → Cohere | Cost window core utilisation | 0.96 | Valid | — |
| Anthropic → Cohere | Cost window successful requests | 264,677 | Valid | — |
| Anthropic → Cohere | Cost window throughput (req/s) | 44,107 | Valid | — |
| Anthropic → Cohere | Cost window threads | 10 | Valid | — |
| Anthropic → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → Cohere | Cost window major page faults | 0 | Valid | — |
| Anthropic → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| Anthropic → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| Anthropic → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Anthropic → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Anthropic → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| Anthropic → Cohere | Streams sustained | — | Untestable | — |
| Anthropic → Cohere | Idle RSS (MiB) | 7.41 | Valid | — |
| Anthropic → Cohere | Steady-state RSS (MiB) | 23.14 | Valid | — |
| Anthropic → Cohere | Recovered RSS (MiB) | 15.3 | Valid | — |
| Anthropic → Cohere | Peak RSS (MiB) | 24.07 | Valid | — |
| Anthropic → Cohere | Peak RSS high-water mark (MiB) | 23.85 | Valid | — |
| Anthropic → Cohere | Time to plateau (s) | — | Not measured | — |
| Anthropic → Cohere | Memory growth (MiB/min) | 0.453 | Valid | — |
| Anthropic → Cohere | Memory load duration (s) | 300 | Valid | — |
| Anthropic → Cohere | Memory shape under load | 0 | Valid | — |
| Anthropic → Cohere | Idle memory shape | 0 | Valid | — |
| Anthropic → Cohere | Idle memory static | 0 | Valid | — |
| Anthropic → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → Cohere | Throughput (req/s) | 47,488 | concurrency 16; n=284,957 | 1 ms |
| Anthropic → Cohere | Throughput (req/s) | 48,799 | concurrency 64; n=292,804 | 5 ms |
| Anthropic → Cohere | Throughput (req/s) | 48,799 | concurrency 64; n=292,804 | 10 ms |
| Anthropic → Cohere | Throughput (req/s) | 48,799 | concurrency 64; n=292,804 | 50 ms |
| Anthropic → Cohere | Throughput (req/s) | 48,799 | concurrency 64; n=292,804 | 100 ms |
| Anthropic → Cohere | Throughput (req/s) | 48,799 | concurrency 64; n=292,804 | Unbounded |
| OpenAI → Cohere | Added latency p50 (µs) | 96 | Valid · n=49,997 gateway / 264,312 direct | — |
| OpenAI → Cohere | Added latency p99 (µs) | 109 | Valid · n=49,997 gateway / 264,312 direct | — |
| OpenAI → Cohere | Gateway p99 at concurrency 1 (µs) | 136 | Valid · n=49,997 | — |
| OpenAI → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=264,312 | — |
| OpenAI → Cohere | CPU per request (µs) | 87.44 | Valid | — |
| OpenAI → Cohere | Requests per CPU-second | 11,436.7 | Valid | — |
| OpenAI → Cohere | Cost window concurrency | 8 | Valid | — |
| OpenAI → Cohere | Cost window core utilisation | 0.963 | Valid | — |
| OpenAI → Cohere | Cost window successful requests | 265,216 | Valid | — |
| OpenAI → Cohere | Cost window throughput (req/s) | 44,197 | Valid | — |
| OpenAI → Cohere | Cost window threads | 10 | Valid | — |
| OpenAI → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → Cohere | Cost window major page faults | 0 | Valid | — |
| OpenAI → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI → Cohere | Streams sustained | — | Untestable | — |
| OpenAI → Cohere | Idle RSS (MiB) | 7.44 | Valid | — |
| OpenAI → Cohere | Steady-state RSS (MiB) | 21.41 | Valid | — |
| OpenAI → Cohere | Recovered RSS (MiB) | 16.35 | Valid | — |
| OpenAI → Cohere | Peak RSS (MiB) | 21.8 | Valid | — |
| OpenAI → Cohere | Peak RSS high-water mark (MiB) | 21.8 | Valid | — |
| OpenAI → Cohere | Time to plateau (s) | — | Not measured | — |
| OpenAI → Cohere | Memory growth (MiB/min) | -0.051 | Valid | — |
| OpenAI → Cohere | Memory load duration (s) | 300 | Valid | — |
| OpenAI → Cohere | Memory shape under load | 0 | Valid | — |
| OpenAI → Cohere | Idle memory shape | 0 | Valid | — |
| OpenAI → Cohere | Idle memory static | 0 | Valid | — |
| OpenAI → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → Cohere | Throughput (req/s) | 47,491 | concurrency 16; n=284,980 | 1 ms |
| OpenAI → Cohere | Throughput (req/s) | 49,018 | concurrency 128; n=294,145 | 5 ms |
| OpenAI → Cohere | Throughput (req/s) | 49,018 | concurrency 128; n=294,145 | 10 ms |
| OpenAI → Cohere | Throughput (req/s) | 49,018 | concurrency 128; n=294,145 | 50 ms |
| OpenAI → Cohere | Throughput (req/s) | 49,018 | concurrency 128; n=294,145 | 100 ms |
| OpenAI → Cohere | Throughput (req/s) | 49,018 | concurrency 128; n=294,145 | Unbounded |
| OpenAI Responses → Cohere | Added latency p50 (µs) | 142 | Valid · n=36,123 gateway / 267,803 direct | — |
| OpenAI Responses → Cohere | Added latency p99 (µs) | 157 | Valid · n=36,123 gateway / 267,803 direct | — |
| OpenAI Responses → Cohere | Gateway p99 at concurrency 1 (µs) | 184 | Valid · n=36,123 | — |
| OpenAI Responses → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=267,803 | — |
| OpenAI Responses → Cohere | CPU per request (µs) | 133.2 | Valid | — |
| OpenAI Responses → Cohere | Requests per CPU-second | 7,506.3 | Valid | — |
| OpenAI Responses → Cohere | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → Cohere | Cost window core utilisation | 0.98 | Valid | — |
| OpenAI Responses → Cohere | Cost window successful requests | 177,224 | Valid | — |
| OpenAI Responses → Cohere | Cost window throughput (req/s) | 29,533 | Valid | — |
| OpenAI Responses → Cohere | Cost window threads | 10 | Valid | — |
| OpenAI Responses → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → Cohere | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → Cohere | Streams sustained | — | Untestable | — |
| OpenAI Responses → Cohere | Idle RSS (MiB) | 7.42 | Valid | — |
| OpenAI Responses → Cohere | Steady-state RSS (MiB) | 23.78 | Valid | — |
| OpenAI Responses → Cohere | Recovered RSS (MiB) | 15.77 | Valid | — |
| OpenAI Responses → Cohere | Peak RSS (MiB) | 24.53 | Valid | — |
| OpenAI Responses → Cohere | Peak RSS high-water mark (MiB) | 24 | Valid | — |
| OpenAI Responses → Cohere | Time to plateau (s) | — | Not measured | — |
| OpenAI Responses → Cohere | Memory growth (MiB/min) | 0.67 | Valid | — |
| OpenAI Responses → Cohere | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → Cohere | Memory shape under load | 1 | Valid | — |
| OpenAI Responses → Cohere | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → Cohere | Idle memory static | 0 | Valid | — |
| OpenAI Responses → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,560 | concurrency 16; n=183,378 | 1 ms |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,675 | concurrency 32; n=184,074 | 5 ms |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,675 | concurrency 32; n=184,074 | 10 ms |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,675 | concurrency 32; n=184,074 | 50 ms |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,675 | concurrency 32; n=184,074 | 100 ms |
| OpenAI Responses → Cohere | Throughput (req/s) | 30,675 | concurrency 32; n=184,074 | Unbounded |
| Gemini → Cohere | Added latency p50 (µs) | 99 | Valid · n=48,806 gateway / 265,602 direct | — |
| Gemini → Cohere | Added latency p99 (µs) | 112 | Valid · n=48,806 gateway / 265,602 direct | — |
| Gemini → Cohere | Gateway p99 at concurrency 1 (µs) | 139 | Valid · n=48,806 | — |
| Gemini → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=265,602 | — |
| Gemini → Cohere | CPU per request (µs) | 91.58 | Valid | — |
| Gemini → Cohere | Requests per CPU-second | 10,919.4 | Valid | — |
| Gemini → Cohere | Cost window concurrency | 8 | Valid | — |
| Gemini → Cohere | Cost window core utilisation | 0.961 | Valid | — |
| Gemini → Cohere | Cost window successful requests | 252,784 | Valid | — |
| Gemini → Cohere | Cost window throughput (req/s) | 42,125 | Valid | — |
| Gemini → Cohere | Cost window threads | 10 | Valid | — |
| Gemini → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → Cohere | Cost window major page faults | 0 | Valid | — |
| Gemini → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → Cohere | Streams sustained | — | Untestable | — |
| Gemini → Cohere | Idle RSS (MiB) | 7.45 | Valid | — |
| Gemini → Cohere | Steady-state RSS (MiB) | 20.93 | Valid | — |
| Gemini → Cohere | Recovered RSS (MiB) | 15.79 | Valid | — |
| Gemini → Cohere | Peak RSS (MiB) | 21.64 | Valid | — |
| Gemini → Cohere | Peak RSS high-water mark (MiB) | 21.36 | Valid | — |
| Gemini → Cohere | Time to plateau (s) | — | Not measured | — |
| Gemini → Cohere | Memory growth (MiB/min) | 0.269 | Valid | — |
| Gemini → Cohere | Memory load duration (s) | 300 | Valid | — |
| Gemini → Cohere | Memory shape under load | 0 | Valid | — |
| Gemini → Cohere | Idle memory shape | 0 | Valid | — |
| Gemini → Cohere | Idle memory static | 0 | Valid | — |
| Gemini → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → Cohere | Throughput (req/s) | 45,456 | concurrency 16; n=272,767 | 1 ms |
| Gemini → Cohere | Throughput (req/s) | 46,559 | concurrency 128; n=279,399 | 5 ms |
| Gemini → Cohere | Throughput (req/s) | 46,559 | concurrency 128; n=279,399 | 10 ms |
| Gemini → Cohere | Throughput (req/s) | 46,559 | concurrency 128; n=279,399 | 50 ms |
| Gemini → Cohere | Throughput (req/s) | 46,559 | concurrency 128; n=279,399 | 100 ms |
| Gemini → Cohere | Throughput (req/s) | 46,559 | concurrency 128; n=279,399 | Unbounded |
| Cohere → Cohere | Added latency p50 (µs) | 78 | Valid · n=59,104 gateway / 267,546 direct | — |
| Cohere → Cohere | Added latency p99 (µs) | 89 | Valid · n=59,104 gateway / 267,546 direct | — |
| Cohere → Cohere | Gateway p99 at concurrency 1 (µs) | 116 | Valid · n=59,104 | — |
| Cohere → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=267,546 | — |
| Cohere → Cohere | CPU per request (µs) | 69.34 | Valid | — |
| Cohere → Cohere | Requests per CPU-second | 14,422.1 | Valid | — |
| Cohere → Cohere | Cost window concurrency | 8 | Valid | — |
| Cohere → Cohere | Cost window core utilisation | 0.957 | Valid | — |
| Cohere → Cohere | Cost window successful requests | 332,285 | Valid | — |
| Cohere → Cohere | Cost window throughput (req/s) | 55,375 | Valid | — |
| Cohere → Cohere | Cost window threads | 10 | Valid | — |
| Cohere → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → Cohere | Cost window major page faults | 0 | Valid | — |
| Cohere → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → Cohere | Streams sustained | — | Untestable | — |
| Cohere → Cohere | Idle RSS (MiB) | 7.44 | Valid | — |
| Cohere → Cohere | Steady-state RSS (MiB) | 21.1 | Valid | — |
| Cohere → Cohere | Recovered RSS (MiB) | 15.61 | Valid | — |
| Cohere → Cohere | Peak RSS (MiB) | 21.71 | Valid | — |
| Cohere → Cohere | Peak RSS high-water mark (MiB) | 21.43 | Valid | — |
| Cohere → Cohere | Time to plateau (s) | — | Not measured | — |
| Cohere → Cohere | Memory growth (MiB/min) | -0.038 | Valid | — |
| Cohere → Cohere | Memory load duration (s) | 300 | Valid | — |
| Cohere → Cohere | Memory shape under load | 0 | Valid | — |
| Cohere → Cohere | Idle memory shape | 0 | Valid | — |
| Cohere → Cohere | Idle memory static | 0 | Valid | — |
| Cohere → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → Cohere | Throughput (req/s) | 61,466 | concurrency 32; n=368,809 | 1 ms |
| Cohere → Cohere | Throughput (req/s) | 62,630 | concurrency 64; n=375,843 | 5 ms |
| Cohere → Cohere | Throughput (req/s) | 62,630 | concurrency 64; n=375,843 | 10 ms |
| Cohere → Cohere | Throughput (req/s) | 62,630 | concurrency 64; n=375,843 | 50 ms |
| Cohere → Cohere | Throughput (req/s) | 62,630 | concurrency 64; n=375,843 | 100 ms |
| Cohere → Cohere | Throughput (req/s) | 62,630 | concurrency 64; n=375,843 | Unbounded |
| Bedrock → Cohere | Added latency p50 (µs) | 103 | Valid · n=47,542 gateway / 266,764 direct | — |
| Bedrock → Cohere | Added latency p99 (µs) | 115 | Valid · n=47,542 gateway / 266,764 direct | — |
| Bedrock → Cohere | Gateway p99 at concurrency 1 (µs) | 142 | Valid · n=47,542 | — |
| Bedrock → Cohere | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=266,764 | — |
| Bedrock → Cohere | CPU per request (µs) | 91.37 | Valid | — |
| Bedrock → Cohere | Requests per CPU-second | 10,944.3 | Valid | — |
| Bedrock → Cohere | Cost window concurrency | 8 | Valid | — |
| Bedrock → Cohere | Cost window core utilisation | 0.96 | Valid | — |
| Bedrock → Cohere | Cost window successful requests | 252,922 | Valid | — |
| Bedrock → Cohere | Cost window throughput (req/s) | 42,148 | Valid | — |
| Bedrock → Cohere | Cost window threads | 10 | Valid | — |
| Bedrock → Cohere | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → Cohere | Cost window major page faults | 0 | Valid | — |
| Bedrock → Cohere | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → Cohere | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → Cohere | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → Cohere | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → Cohere | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → Cohere | Streams sustained | — | Untestable | — |
| Bedrock → Cohere | Idle RSS (MiB) | 7.43 | Valid | — |
| Bedrock → Cohere | Steady-state RSS (MiB) | 22.61 | Valid | — |
| Bedrock → Cohere | Recovered RSS (MiB) | 15.53 | Valid | — |
| Bedrock → Cohere | Peak RSS (MiB) | 23.04 | Valid | — |
| Bedrock → Cohere | Peak RSS high-water mark (MiB) | 22.74 | Valid | — |
| Bedrock → Cohere | Time to plateau (s) | — | Not measured | — |
| Bedrock → Cohere | Memory growth (MiB/min) | 0.056 | Valid | — |
| Bedrock → Cohere | Memory load duration (s) | 300 | Valid | — |
| Bedrock → Cohere | Memory shape under load | 0 | Valid | — |
| Bedrock → Cohere | Idle memory shape | 0 | Valid | — |
| Bedrock → Cohere | Idle memory static | 0 | Valid | — |
| Bedrock → Cohere | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → Cohere | Throughput (req/s) | 45,878 | concurrency 16; n=275,299 | 1 ms |
| Bedrock → Cohere | Throughput (req/s) | 46,932 | concurrency 64; n=281,601 | 5 ms |
| Bedrock → Cohere | Throughput (req/s) | 46,932 | concurrency 64; n=281,601 | 10 ms |
| Bedrock → Cohere | Throughput (req/s) | 46,932 | concurrency 64; n=281,601 | 50 ms |
| Bedrock → Cohere | Throughput (req/s) | 46,932 | concurrency 64; n=281,601 | 100 ms |
| Bedrock → Cohere | Throughput (req/s) | 46,932 | concurrency 64; n=281,601 | Unbounded |
| Gemini → OpenAI Responses | Added latency p50 (µs) | 108 | Valid · n=45,574 gateway / 262,793 direct | — |
| Gemini → OpenAI Responses | Added latency p99 (µs) | 121 | Valid · n=45,574 gateway / 262,793 direct | — |
| Gemini → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 149 | Valid · n=45,574 | — |
| Gemini → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=262,793 | — |
| Gemini → OpenAI Responses | CPU per request (µs) | 96.57 | Valid | — |
| Gemini → OpenAI Responses | Requests per CPU-second | 10,355.5 | Valid | — |
| Gemini → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| Gemini → OpenAI Responses | Cost window core utilisation | 0.963 | Valid | — |
| Gemini → OpenAI Responses | Cost window successful requests | 240,144 | Valid | — |
| Gemini → OpenAI Responses | Cost window throughput (req/s) | 40,019 | Valid | — |
| Gemini → OpenAI Responses | Cost window threads | 10 | Valid | — |
| Gemini → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| Gemini → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → OpenAI Responses | Streams sustained | — | Untestable | — |
| Gemini → OpenAI Responses | Idle RSS (MiB) | 7.43 | Valid | — |
| Gemini → OpenAI Responses | Steady-state RSS (MiB) | 22.65 | Valid | — |
| Gemini → OpenAI Responses | Recovered RSS (MiB) | 15.47 | Valid | — |
| Gemini → OpenAI Responses | Peak RSS (MiB) | 23.53 | Valid | — |
| Gemini → OpenAI Responses | Peak RSS high-water mark (MiB) | 23.53 | Valid | — |
| Gemini → OpenAI Responses | Time to plateau (s) | — | Not measured | — |
| Gemini → OpenAI Responses | Memory growth (MiB/min) | 1.35 | Valid | — |
| Gemini → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| Gemini → OpenAI Responses | Memory shape under load | 1 | Valid | — |
| Gemini → OpenAI Responses | Idle memory shape | 0 | Valid | — |
| Gemini → OpenAI Responses | Idle memory static | 0 | Valid | — |
| Gemini → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → OpenAI Responses | Throughput (req/s) | 42,669 | concurrency 16; n=256,046 | 1 ms |
| Gemini → OpenAI Responses | Throughput (req/s) | 43,426 | concurrency 64; n=260,577 | 5 ms |
| Gemini → OpenAI Responses | Throughput (req/s) | 43,426 | concurrency 64; n=260,577 | 10 ms |
| Gemini → OpenAI Responses | Throughput (req/s) | 43,426 | concurrency 64; n=260,577 | 50 ms |
| Gemini → OpenAI Responses | Throughput (req/s) | 43,426 | concurrency 64; n=260,577 | 100 ms |
| Gemini → OpenAI Responses | Throughput (req/s) | 43,426 | concurrency 64; n=260,577 | Unbounded |
| Anthropic → OpenAI Responses | Added latency p50 (µs) | 105 | Valid · n=46,499 gateway / 259,447 direct | — |
| Anthropic → OpenAI Responses | Added latency p99 (µs) | 118 | Valid · n=46,499 gateway / 259,447 direct | — |
| Anthropic → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 146 | Valid · n=46,499 | — |
| Anthropic → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=259,447 | — |
| Anthropic → OpenAI Responses | CPU per request (µs) | 93.29 | Valid | — |
| Anthropic → OpenAI Responses | Requests per CPU-second | 10,719.4 | Valid | — |
| Anthropic → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| Anthropic → OpenAI Responses | Cost window core utilisation | 0.961 | Valid | — |
| Anthropic → OpenAI Responses | Cost window successful requests | 248,048 | Valid | — |
| Anthropic → OpenAI Responses | Cost window throughput (req/s) | 41,335 | Valid | — |
| Anthropic → OpenAI Responses | Cost window threads | 10 | Valid | — |
| Anthropic → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| Anthropic → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| Anthropic → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| Anthropic → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Anthropic → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Anthropic → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| Anthropic → OpenAI Responses | Streams sustained | — | Untestable | — |
| Anthropic → OpenAI Responses | Idle RSS (MiB) | 7.42 | Valid | — |
| Anthropic → OpenAI Responses | Steady-state RSS (MiB) | 21.11 | Valid | — |
| Anthropic → OpenAI Responses | Recovered RSS (MiB) | 15.63 | Valid | — |
| Anthropic → OpenAI Responses | Peak RSS (MiB) | 21.59 | Valid | — |
| Anthropic → OpenAI Responses | Peak RSS high-water mark (MiB) | 21.47 | Valid | — |
| Anthropic → OpenAI Responses | Time to plateau (s) | 156 | Valid | — |
| Anthropic → OpenAI Responses | Memory growth (MiB/min) | -0.108 | Valid | — |
| Anthropic → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| Anthropic → OpenAI Responses | Memory shape under load | — | Not measured | — |
| Anthropic → OpenAI Responses | Idle memory shape | 0 | Valid | — |
| Anthropic → OpenAI Responses | Idle memory static | 0 | Valid | — |
| Anthropic → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → OpenAI Responses | Throughput (req/s) | 44,406 | concurrency 16; n=266,467 | 1 ms |
| Anthropic → OpenAI Responses | Throughput (req/s) | 45,442 | concurrency 64; n=272,657 | 5 ms |
| Anthropic → OpenAI Responses | Throughput (req/s) | 45,442 | concurrency 64; n=272,657 | 10 ms |
| Anthropic → OpenAI Responses | Throughput (req/s) | 45,442 | concurrency 64; n=272,657 | 50 ms |
| Anthropic → OpenAI Responses | Throughput (req/s) | 45,442 | concurrency 64; n=272,657 | 100 ms |
| Anthropic → OpenAI Responses | Throughput (req/s) | 45,442 | concurrency 64; n=272,657 | Unbounded |
| OpenAI → OpenAI Responses | Added latency p50 (µs) | 104 | Valid · n=46,929 gateway / 263,227 direct | — |
| OpenAI → OpenAI Responses | Added latency p99 (µs) | 117 | Valid · n=46,929 gateway / 263,227 direct | — |
| OpenAI → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 145 | Valid · n=46,929 | — |
| OpenAI → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=263,227 | — |
| OpenAI → OpenAI Responses | CPU per request (µs) | 92.93 | Valid | — |
| OpenAI → OpenAI Responses | Requests per CPU-second | 10,760.5 | Valid | — |
| OpenAI → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| OpenAI → OpenAI Responses | Cost window core utilisation | 0.961 | Valid | — |
| OpenAI → OpenAI Responses | Cost window successful requests | 248,997 | Valid | — |
| OpenAI → OpenAI Responses | Cost window throughput (req/s) | 41,494 | Valid | — |
| OpenAI → OpenAI Responses | Cost window threads | 10 | Valid | — |
| OpenAI → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| OpenAI → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI → OpenAI Responses | Streams sustained | — | Untestable | — |
| OpenAI → OpenAI Responses | Idle RSS (MiB) | 7.41 | Valid | — |
| OpenAI → OpenAI Responses | Steady-state RSS (MiB) | 21.27 | Valid | — |
| OpenAI → OpenAI Responses | Recovered RSS (MiB) | 15.77 | Valid | — |
| OpenAI → OpenAI Responses | Peak RSS (MiB) | 21.65 | Valid | — |
| OpenAI → OpenAI Responses | Peak RSS high-water mark (MiB) | 21.43 | Valid | — |
| OpenAI → OpenAI Responses | Time to plateau (s) | 246 | Valid | — |
| OpenAI → OpenAI Responses | Memory growth (MiB/min) | 0.069 | Valid | — |
| OpenAI → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| OpenAI → OpenAI Responses | Memory shape under load | — | Not measured | — |
| OpenAI → OpenAI Responses | Idle memory shape | — | Not measured | — |
| OpenAI → OpenAI Responses | Idle memory static | 1 | Valid | — |
| OpenAI → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → OpenAI Responses | Throughput (req/s) | 44,763 | concurrency 16; n=268,603 | 1 ms |
| OpenAI → OpenAI Responses | Throughput (req/s) | 45,770 | concurrency 64; n=274,629 | 5 ms |
| OpenAI → OpenAI Responses | Throughput (req/s) | 45,770 | concurrency 64; n=274,629 | 10 ms |
| OpenAI → OpenAI Responses | Throughput (req/s) | 45,770 | concurrency 64; n=274,629 | 50 ms |
| OpenAI → OpenAI Responses | Throughput (req/s) | 45,770 | concurrency 64; n=274,629 | 100 ms |
| OpenAI → OpenAI Responses | Throughput (req/s) | 45,770 | concurrency 64; n=274,629 | Unbounded |
| Cohere → OpenAI Responses | Added latency p50 (µs) | 102 | Valid · n=47,752 gateway / 268,640 direct | — |
| Cohere → OpenAI Responses | Added latency p99 (µs) | 116 | Valid · n=47,752 gateway / 268,640 direct | — |
| Cohere → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 143 | Valid · n=47,752 | — |
| Cohere → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=268,640 | — |
| Cohere → OpenAI Responses | CPU per request (µs) | 91.48 | Valid | — |
| Cohere → OpenAI Responses | Requests per CPU-second | 10,931.6 | Valid | — |
| Cohere → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| Cohere → OpenAI Responses | Cost window core utilisation | 0.96 | Valid | — |
| Cohere → OpenAI Responses | Cost window successful requests | 252,738 | Valid | — |
| Cohere → OpenAI Responses | Cost window throughput (req/s) | 42,119 | Valid | — |
| Cohere → OpenAI Responses | Cost window threads | 10 | Valid | — |
| Cohere → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| Cohere → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → OpenAI Responses | Streams sustained | — | Untestable | — |
| Cohere → OpenAI Responses | Idle RSS (MiB) | 7.44 | Valid | — |
| Cohere → OpenAI Responses | Steady-state RSS (MiB) | 21.58 | Valid | — |
| Cohere → OpenAI Responses | Recovered RSS (MiB) | 16.15 | Valid | — |
| Cohere → OpenAI Responses | Peak RSS (MiB) | 22.04 | Valid | — |
| Cohere → OpenAI Responses | Peak RSS high-water mark (MiB) | 21.62 | Valid | — |
| Cohere → OpenAI Responses | Time to plateau (s) | — | Not measured | — |
| Cohere → OpenAI Responses | Memory growth (MiB/min) | -0.474 | Valid | — |
| Cohere → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| Cohere → OpenAI Responses | Memory shape under load | -1 | Valid | — |
| Cohere → OpenAI Responses | Idle memory shape | 0 | Valid | — |
| Cohere → OpenAI Responses | Idle memory static | 0 | Valid | — |
| Cohere → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → OpenAI Responses | Throughput (req/s) | 45,427 | concurrency 16; n=272,592 | 1 ms |
| Cohere → OpenAI Responses | Throughput (req/s) | 46,419 | concurrency 64; n=278,528 | 5 ms |
| Cohere → OpenAI Responses | Throughput (req/s) | 46,419 | concurrency 64; n=278,528 | 10 ms |
| Cohere → OpenAI Responses | Throughput (req/s) | 46,419 | concurrency 64; n=278,528 | 50 ms |
| Cohere → OpenAI Responses | Throughput (req/s) | 46,419 | concurrency 64; n=278,528 | 100 ms |
| Cohere → OpenAI Responses | Throughput (req/s) | 46,419 | concurrency 64; n=278,528 | Unbounded |
| OpenAI Responses → OpenAI Responses | Added latency p50 (µs) | 79 | Valid · n=58,401 gateway / 264,340 direct | — |
| OpenAI Responses → OpenAI Responses | Added latency p99 (µs) | 90 | Valid · n=58,401 gateway / 264,340 direct | — |
| OpenAI Responses → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 118 | Valid · n=58,401 | — |
| OpenAI Responses → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=264,340 | — |
| OpenAI Responses → OpenAI Responses | CPU per request (µs) | 70.64 | Valid | — |
| OpenAI Responses → OpenAI Responses | Requests per CPU-second | 14,157 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window core utilisation | 0.956 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window successful requests | 326,035 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window throughput (req/s) | 54,332 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window threads | 10 | Valid | — |
| OpenAI Responses → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Streams sustained | — | Untestable | — |
| OpenAI Responses → OpenAI Responses | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI Responses → OpenAI Responses | Steady-state RSS (MiB) | 20.77 | Valid | — |
| OpenAI Responses → OpenAI Responses | Recovered RSS (MiB) | 15.61 | Valid | — |
| OpenAI Responses → OpenAI Responses | Peak RSS (MiB) | 21.49 | Valid | — |
| OpenAI Responses → OpenAI Responses | Peak RSS high-water mark (MiB) | 21.43 | Valid | — |
| OpenAI Responses → OpenAI Responses | Time to plateau (s) | — | Not measured | — |
| OpenAI Responses → OpenAI Responses | Memory growth (MiB/min) | 0.682 | Valid | — |
| OpenAI Responses → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → OpenAI Responses | Memory shape under load | 1 | Valid | — |
| OpenAI Responses → OpenAI Responses | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → OpenAI Responses | Idle memory static | 0 | Valid | — |
| OpenAI Responses → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 59,818 | concurrency 32; n=358,944 | 1 ms |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 60,832 | concurrency 128; n=365,044 | 5 ms |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 60,832 | concurrency 128; n=365,044 | 10 ms |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 60,832 | concurrency 128; n=365,044 | 50 ms |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 60,832 | concurrency 128; n=365,044 | 100 ms |
| OpenAI Responses → OpenAI Responses | Throughput (req/s) | 60,832 | concurrency 128; n=365,044 | Unbounded |
| Bedrock → OpenAI Responses | Added latency p50 (µs) | 110 | Valid · n=44,818 gateway / 263,285 direct | — |
| Bedrock → OpenAI Responses | Added latency p99 (µs) | 123 | Valid · n=44,818 gateway / 263,285 direct | — |
| Bedrock → OpenAI Responses | Gateway p99 at concurrency 1 (µs) | 151 | Valid · n=44,818 | — |
| Bedrock → OpenAI Responses | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=263,285 | — |
| Bedrock → OpenAI Responses | CPU per request (µs) | 97.04 | Valid | — |
| Bedrock → OpenAI Responses | Requests per CPU-second | 10,305.5 | Valid | — |
| Bedrock → OpenAI Responses | Cost window concurrency | 8 | Valid | — |
| Bedrock → OpenAI Responses | Cost window core utilisation | 0.96 | Valid | — |
| Bedrock → OpenAI Responses | Cost window successful requests | 238,263 | Valid | — |
| Bedrock → OpenAI Responses | Cost window throughput (req/s) | 39,706 | Valid | — |
| Bedrock → OpenAI Responses | Cost window threads | 10 | Valid | — |
| Bedrock → OpenAI Responses | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → OpenAI Responses | Cost window major page faults | 0 | Valid | — |
| Bedrock → OpenAI Responses | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → OpenAI Responses | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → OpenAI Responses | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → OpenAI Responses | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → OpenAI Responses | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → OpenAI Responses | Streams sustained | — | Untestable | — |
| Bedrock → OpenAI Responses | Idle RSS (MiB) | 7.43 | Valid | — |
| Bedrock → OpenAI Responses | Steady-state RSS (MiB) | 22.62 | Valid | — |
| Bedrock → OpenAI Responses | Recovered RSS (MiB) | 15.99 | Valid | — |
| Bedrock → OpenAI Responses | Peak RSS (MiB) | 23.49 | Valid | — |
| Bedrock → OpenAI Responses | Peak RSS high-water mark (MiB) | 23.33 | Valid | — |
| Bedrock → OpenAI Responses | Time to plateau (s) | — | Not measured | — |
| Bedrock → OpenAI Responses | Memory growth (MiB/min) | 0.59 | Valid | — |
| Bedrock → OpenAI Responses | Memory load duration (s) | 300 | Valid | — |
| Bedrock → OpenAI Responses | Memory shape under load | 1 | Valid | — |
| Bedrock → OpenAI Responses | Idle memory shape | 0 | Valid | — |
| Bedrock → OpenAI Responses | Idle memory static | 0 | Valid | — |
| Bedrock → OpenAI Responses | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → OpenAI Responses | Throughput (req/s) | 42,923 | concurrency 16; n=257,569 | 1 ms |
| Bedrock → OpenAI Responses | Throughput (req/s) | 43,797 | concurrency 64; n=262,796 | 5 ms |
| Bedrock → OpenAI Responses | Throughput (req/s) | 43,797 | concurrency 64; n=262,796 | 10 ms |
| Bedrock → OpenAI Responses | Throughput (req/s) | 43,797 | concurrency 64; n=262,796 | 50 ms |
| Bedrock → OpenAI Responses | Throughput (req/s) | 43,797 | concurrency 64; n=262,796 | 100 ms |
| Bedrock → OpenAI Responses | Throughput (req/s) | 43,797 | concurrency 64; n=262,796 | Unbounded |
| Anthropic → Gemini | Added latency p50 (µs) | 101 | Valid · n=48,133 gateway / 261,686 direct | — |
| Anthropic → Gemini | Added latency p99 (µs) | 113 | Valid · n=48,133 gateway / 261,686 direct | — |
| Anthropic → Gemini | Gateway p99 at concurrency 1 (µs) | 141 | Valid · n=48,133 | — |
| Anthropic → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=261,686 | — |
| Anthropic → Gemini | CPU per request (µs) | 90.32 | Valid | — |
| Anthropic → Gemini | Requests per CPU-second | 11,071.3 | Valid | — |
| Anthropic → Gemini | Cost window concurrency | 8 | Valid | — |
| Anthropic → Gemini | Cost window core utilisation | 0.964 | Valid | — |
| Anthropic → Gemini | Cost window successful requests | 256,964 | Valid | — |
| Anthropic → Gemini | Cost window throughput (req/s) | 42,822 | Valid | — |
| Anthropic → Gemini | Cost window threads | 10 | Valid | — |
| Anthropic → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| Anthropic → Gemini | Cost window major page faults | 0 | Valid | — |
| Anthropic → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| Anthropic → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| Anthropic → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Anthropic → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Anthropic → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| Anthropic → Gemini | Streams sustained | — | Untestable | — |
| Anthropic → Gemini | Idle RSS (MiB) | 7.44 | Valid | — |
| Anthropic → Gemini | Steady-state RSS (MiB) | 20.82 | Valid | — |
| Anthropic → Gemini | Recovered RSS (MiB) | 15.73 | Valid | — |
| Anthropic → Gemini | Peak RSS (MiB) | 21.05 | Valid | — |
| Anthropic → Gemini | Peak RSS high-water mark (MiB) | 20.84 | Valid | — |
| Anthropic → Gemini | Time to plateau (s) | — | Not measured | — |
| Anthropic → Gemini | Memory growth (MiB/min) | 0.347 | Valid | — |
| Anthropic → Gemini | Memory load duration (s) | 300 | Valid | — |
| Anthropic → Gemini | Memory shape under load | 0 | Valid | — |
| Anthropic → Gemini | Idle memory shape | 0 | Valid | — |
| Anthropic → Gemini | Idle memory static | 0 | Valid | — |
| Anthropic → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Anthropic → Gemini | Throughput (req/s) | 45,587 | concurrency 16; n=273,548 | 1 ms |
| Anthropic → Gemini | Throughput (req/s) | 46,654 | concurrency 64; n=279,944 | 5 ms |
| Anthropic → Gemini | Throughput (req/s) | 46,654 | concurrency 64; n=279,944 | 10 ms |
| Anthropic → Gemini | Throughput (req/s) | 46,654 | concurrency 64; n=279,944 | 50 ms |
| Anthropic → Gemini | Throughput (req/s) | 46,654 | concurrency 64; n=279,944 | 100 ms |
| Anthropic → Gemini | Throughput (req/s) | 46,654 | concurrency 64; n=279,944 | Unbounded |
| Bedrock → Gemini | Added latency p50 (µs) | 104 | Valid · n=47,051 gateway / 267,236 direct | — |
| Bedrock → Gemini | Added latency p99 (µs) | 117 | Valid · n=47,051 gateway / 267,236 direct | — |
| Bedrock → Gemini | Gateway p99 at concurrency 1 (µs) | 144 | Valid · n=47,051 | — |
| Bedrock → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=267,236 | — |
| Bedrock → Gemini | CPU per request (µs) | 93.43 | Valid | — |
| Bedrock → Gemini | Requests per CPU-second | 10,702.9 | Valid | — |
| Bedrock → Gemini | Cost window concurrency | 8 | Valid | — |
| Bedrock → Gemini | Cost window core utilisation | 0.96 | Valid | — |
| Bedrock → Gemini | Cost window successful requests | 247,451 | Valid | — |
| Bedrock → Gemini | Cost window throughput (req/s) | 41,236 | Valid | — |
| Bedrock → Gemini | Cost window threads | 10 | Valid | — |
| Bedrock → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| Bedrock → Gemini | Cost window major page faults | 0 | Valid | — |
| Bedrock → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| Bedrock → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| Bedrock → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Bedrock → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Bedrock → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| Bedrock → Gemini | Streams sustained | — | Untestable | — |
| Bedrock → Gemini | Idle RSS (MiB) | 7.42 | Valid | — |
| Bedrock → Gemini | Steady-state RSS (MiB) | 21.51 | Valid | — |
| Bedrock → Gemini | Recovered RSS (MiB) | 15.64 | Valid | — |
| Bedrock → Gemini | Peak RSS (MiB) | 22.76 | Valid | — |
| Bedrock → Gemini | Peak RSS high-water mark (MiB) | 22.25 | Valid | — |
| Bedrock → Gemini | Time to plateau (s) | — | Not measured | — |
| Bedrock → Gemini | Memory growth (MiB/min) | 0.651 | Valid | — |
| Bedrock → Gemini | Memory load duration (s) | 300 | Valid | — |
| Bedrock → Gemini | Memory shape under load | 1 | Valid | — |
| Bedrock → Gemini | Idle memory shape | 0 | Valid | — |
| Bedrock → Gemini | Idle memory static | 0 | Valid | — |
| Bedrock → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Bedrock → Gemini | Throughput (req/s) | 44,045 | concurrency 16; n=264,301 | 1 ms |
| Bedrock → Gemini | Throughput (req/s) | 45,269 | concurrency 64; n=271,622 | 5 ms |
| Bedrock → Gemini | Throughput (req/s) | 45,269 | concurrency 64; n=271,622 | 10 ms |
| Bedrock → Gemini | Throughput (req/s) | 45,269 | concurrency 64; n=271,622 | 50 ms |
| Bedrock → Gemini | Throughput (req/s) | 45,269 | concurrency 64; n=271,622 | 100 ms |
| Bedrock → Gemini | Throughput (req/s) | 45,269 | concurrency 64; n=271,622 | Unbounded |
| Gemini → Gemini | Added latency p50 (µs) | 87 | Valid · n=54,333 gateway / 264,586 direct | — |
| Gemini → Gemini | Added latency p99 (µs) | 97 | Valid · n=54,333 gateway / 264,586 direct | — |
| Gemini → Gemini | Gateway p99 at concurrency 1 (µs) | 125 | Valid · n=54,333 | — |
| Gemini → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=264,586 | — |
| Gemini → Gemini | CPU per request (µs) | 77.73 | Valid | — |
| Gemini → Gemini | Requests per CPU-second | 12,864.3 | Valid | — |
| Gemini → Gemini | Cost window concurrency | 8 | Valid | — |
| Gemini → Gemini | Cost window core utilisation | 0.962 | Valid | — |
| Gemini → Gemini | Cost window successful requests | 298,065 | Valid | — |
| Gemini → Gemini | Cost window throughput (req/s) | 49,672 | Valid | — |
| Gemini → Gemini | Cost window threads | 10 | Valid | — |
| Gemini → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| Gemini → Gemini | Cost window major page faults | 0 | Valid | — |
| Gemini → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| Gemini → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| Gemini → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Gemini → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Gemini → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| Gemini → Gemini | Streams sustained | — | Untestable | — |
| Gemini → Gemini | Idle RSS (MiB) | 7.42 | Valid | — |
| Gemini → Gemini | Steady-state RSS (MiB) | 21.42 | Valid | — |
| Gemini → Gemini | Recovered RSS (MiB) | 15.91 | Valid | — |
| Gemini → Gemini | Peak RSS (MiB) | 21.93 | Valid | — |
| Gemini → Gemini | Peak RSS high-water mark (MiB) | 21.64 | Valid | — |
| Gemini → Gemini | Time to plateau (s) | — | Not measured | — |
| Gemini → Gemini | Memory growth (MiB/min) | 0.002 | Valid | — |
| Gemini → Gemini | Memory load duration (s) | 300 | Valid | — |
| Gemini → Gemini | Memory shape under load | 0 | Valid | — |
| Gemini → Gemini | Idle memory shape | 0 | Valid | — |
| Gemini → Gemini | Idle memory static | 0 | Valid | — |
| Gemini → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Gemini → Gemini | Throughput (req/s) | 54,786 | concurrency 32; n=328,745 | 1 ms |
| Gemini → Gemini | Throughput (req/s) | 55,374 | concurrency 64; n=332,269 | 5 ms |
| Gemini → Gemini | Throughput (req/s) | 55,374 | concurrency 64; n=332,269 | 10 ms |
| Gemini → Gemini | Throughput (req/s) | 55,374 | concurrency 64; n=332,269 | 50 ms |
| Gemini → Gemini | Throughput (req/s) | 55,374 | concurrency 64; n=332,269 | 100 ms |
| Gemini → Gemini | Throughput (req/s) | 55,374 | concurrency 64; n=332,269 | Unbounded |
| OpenAI → Gemini | Added latency p50 (µs) | 99 | Valid · n=48,998 gateway / 264,583 direct | — |
| OpenAI → Gemini | Added latency p99 (µs) | 111 | Valid · n=48,998 gateway / 264,583 direct | — |
| OpenAI → Gemini | Gateway p99 at concurrency 1 (µs) | 138 | Valid · n=48,998 | — |
| OpenAI → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=264,583 | — |
| OpenAI → Gemini | CPU per request (µs) | 89.82 | Valid | — |
| OpenAI → Gemini | Requests per CPU-second | 11,133.8 | Valid | — |
| OpenAI → Gemini | Cost window concurrency | 8 | Valid | — |
| OpenAI → Gemini | Cost window core utilisation | 0.963 | Valid | — |
| OpenAI → Gemini | Cost window successful requests | 258,193 | Valid | — |
| OpenAI → Gemini | Cost window throughput (req/s) | 43,027 | Valid | — |
| OpenAI → Gemini | Cost window threads | 10 | Valid | — |
| OpenAI → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI → Gemini | Cost window major page faults | 0 | Valid | — |
| OpenAI → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI → Gemini | Streams sustained | — | Untestable | — |
| OpenAI → Gemini | Idle RSS (MiB) | 7.42 | Valid | — |
| OpenAI → Gemini | Steady-state RSS (MiB) | 22.4 | Valid | — |
| OpenAI → Gemini | Recovered RSS (MiB) | 14.95 | Valid | — |
| OpenAI → Gemini | Peak RSS (MiB) | 22.73 | Valid | — |
| OpenAI → Gemini | Peak RSS high-water mark (MiB) | 22.73 | Valid | — |
| OpenAI → Gemini | Time to plateau (s) | — | Not measured | — |
| OpenAI → Gemini | Memory growth (MiB/min) | 0.078 | Valid | — |
| OpenAI → Gemini | Memory load duration (s) | 300 | Valid | — |
| OpenAI → Gemini | Memory shape under load | 0 | Valid | — |
| OpenAI → Gemini | Idle memory shape | 0 | Valid | — |
| OpenAI → Gemini | Idle memory static | 0 | Valid | — |
| OpenAI → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI → Gemini | Throughput (req/s) | 46,221 | concurrency 16; n=277,357 | 1 ms |
| OpenAI → Gemini | Throughput (req/s) | 47,238 | concurrency 64; n=283,442 | 5 ms |
| OpenAI → Gemini | Throughput (req/s) | 47,238 | concurrency 64; n=283,442 | 10 ms |
| OpenAI → Gemini | Throughput (req/s) | 47,238 | concurrency 64; n=283,442 | 50 ms |
| OpenAI → Gemini | Throughput (req/s) | 47,238 | concurrency 64; n=283,442 | 100 ms |
| OpenAI → Gemini | Throughput (req/s) | 47,238 | concurrency 64; n=283,442 | Unbounded |
| OpenAI Responses → Gemini | Added latency p50 (µs) | 148 | Valid · n=34,929 gateway / 262,960 direct | — |
| OpenAI Responses → Gemini | Added latency p99 (µs) | 162 | Valid · n=34,929 gateway / 262,960 direct | — |
| OpenAI Responses → Gemini | Gateway p99 at concurrency 1 (µs) | 190 | Valid · n=34,929 | — |
| OpenAI Responses → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 28 | Valid · n=262,960 | — |
| OpenAI Responses → Gemini | CPU per request (µs) | 136.8 | Valid | — |
| OpenAI Responses → Gemini | Requests per CPU-second | 7,308.8 | Valid | — |
| OpenAI Responses → Gemini | Cost window concurrency | 8 | Valid | — |
| OpenAI Responses → Gemini | Cost window core utilisation | 0.981 | Valid | — |
| OpenAI Responses → Gemini | Cost window successful requests | 172,561 | Valid | — |
| OpenAI Responses → Gemini | Cost window throughput (req/s) | 28,757 | Valid | — |
| OpenAI Responses → Gemini | Cost window threads | 10 | Valid | — |
| OpenAI Responses → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| OpenAI Responses → Gemini | Cost window major page faults | 0 | Valid | — |
| OpenAI Responses → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| OpenAI Responses → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| OpenAI Responses → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| OpenAI Responses → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| OpenAI Responses → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| OpenAI Responses → Gemini | Streams sustained | — | Untestable | — |
| OpenAI Responses → Gemini | Idle RSS (MiB) | 7.43 | Valid | — |
| OpenAI Responses → Gemini | Steady-state RSS (MiB) | 21.73 | Valid | — |
| OpenAI Responses → Gemini | Recovered RSS (MiB) | 15.08 | Valid | — |
| OpenAI Responses → Gemini | Peak RSS (MiB) | 22.41 | Valid | — |
| OpenAI Responses → Gemini | Peak RSS high-water mark (MiB) | 21.99 | Valid | — |
| OpenAI Responses → Gemini | Time to plateau (s) | — | Not measured | — |
| OpenAI Responses → Gemini | Memory growth (MiB/min) | 0.429 | Valid | — |
| OpenAI Responses → Gemini | Memory load duration (s) | 300 | Valid | — |
| OpenAI Responses → Gemini | Memory shape under load | 1 | Valid | — |
| OpenAI Responses → Gemini | Idle memory shape | 0 | Valid | — |
| OpenAI Responses → Gemini | Idle memory static | 0 | Valid | — |
| OpenAI Responses → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| OpenAI Responses → Gemini | Throughput (req/s) | 29,952 | concurrency 16; n=179,734 | 1 ms |
| OpenAI Responses → Gemini | Throughput (req/s) | 30,107 | concurrency 32; n=180,664 | 5 ms |
| OpenAI Responses → Gemini | Throughput (req/s) | 30,107 | concurrency 32; n=180,664 | 10 ms |
| OpenAI Responses → Gemini | Throughput (req/s) | 30,107 | concurrency 32; n=180,664 | 50 ms |
| OpenAI Responses → Gemini | Throughput (req/s) | 30,107 | concurrency 32; n=180,664 | 100 ms |
| OpenAI Responses → Gemini | Throughput (req/s) | 30,107 | concurrency 32; n=180,664 | Unbounded |
| Cohere → Gemini | Added latency p50 (µs) | 100 | Valid · n=49,012 gateway / 269,799 direct | — |
| Cohere → Gemini | Added latency p99 (µs) | 112 | Valid · n=49,012 gateway / 269,799 direct | — |
| Cohere → Gemini | Gateway p99 at concurrency 1 (µs) | 139 | Valid · n=49,012 | — |
| Cohere → Gemini | Direct-to-mock p99 at concurrency 1 (µs) | 27 | Valid · n=269,799 | — |
| Cohere → Gemini | CPU per request (µs) | 88.31 | Valid | — |
| Cohere → Gemini | Requests per CPU-second | 11,323.5 | Valid | — |
| Cohere → Gemini | Cost window concurrency | 8 | Valid | — |
| Cohere → Gemini | Cost window core utilisation | 0.961 | Valid | — |
| Cohere → Gemini | Cost window successful requests | 262,140 | Valid | — |
| Cohere → Gemini | Cost window throughput (req/s) | 43,684 | Valid | — |
| Cohere → Gemini | Cost window threads | 10 | Valid | — |
| Cohere → Gemini | Non-voluntary context switches per request | 0 | Valid | — |
| Cohere → Gemini | Cost window major page faults | 0 | Valid | — |
| Cohere → Gemini | Added time to first event p50 (µs) | — | Untestable | — |
| Cohere → Gemini | Added time to first event p99 (µs) | — | Untestable | — |
| Cohere → Gemini | Added inter-frame gap p50 (µs) | — | Untestable | — |
| Cohere → Gemini | Added inter-frame gap p99 (µs) | — | Untestable | — |
| Cohere → Gemini | Streams sustained (frames/s) | — | Untestable | — |
| Cohere → Gemini | Streams sustained | — | Untestable | — |
| Cohere → Gemini | Idle RSS (MiB) | 7.42 | Valid | — |
| Cohere → Gemini | Steady-state RSS (MiB) | 20.26 | Valid | — |
| Cohere → Gemini | Recovered RSS (MiB) | 15.55 | Valid | — |
| Cohere → Gemini | Peak RSS (MiB) | 20.91 | Valid | — |
| Cohere → Gemini | Peak RSS high-water mark (MiB) | 20.59 | Valid | — |
| Cohere → Gemini | Time to plateau (s) | — | Not measured | — |
| Cohere → Gemini | Memory growth (MiB/min) | 0.71 | Valid | — |
| Cohere → Gemini | Memory load duration (s) | 300 | Valid | — |
| Cohere → Gemini | Memory shape under load | 1 | Valid | — |
| Cohere → Gemini | Idle memory shape | 0 | Valid | — |
| Cohere → Gemini | Idle memory static | 0 | Valid | — |
| Cohere → Gemini | Idle memory growth (MiB/min) | 0.004 | Valid | — |
| Cohere → Gemini | Throughput (req/s) | 46,982 | concurrency 16; n=281,924 | 1 ms |
| Cohere → Gemini | Throughput (req/s) | 47,826 | concurrency 64; n=286,965 | 5 ms |
| Cohere → Gemini | Throughput (req/s) | 47,826 | concurrency 64; n=286,965 | 10 ms |
| Cohere → Gemini | Throughput (req/s) | 47,826 | concurrency 64; n=286,965 | 50 ms |
| Cohere → Gemini | Throughput (req/s) | 47,826 | concurrency 64; n=286,965 | 100 ms |
| Cohere → Gemini | Throughput (req/s) | 47,826 | concurrency 64; n=286,965 | Unbounded |
Probe evidence and instrument provenance
Historical capability notes may reference an earlier product version. They are not a new probe or a current support claim. Full RSS samples remain in the immutable raw JSON.
| Ingress → upstream | Cell state | Probe HTTP status | Recorded reason | Recorded detail | Evidence scope | Response excerpt | Tested build |
|---|---|---|---|---|---|---|---|
| Cohere → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → Anthropic | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Cohere → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → Bedrock | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Cohere → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → OpenAI | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Cohere → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → Cohere | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Cohere → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → OpenAI Responses | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Anthropic → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Bedrock → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Gemini → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| OpenAI Responses → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
| Cohere → Gemini | Served | — | — | — | Recorded run evidence | — | getbusbar/busbar:1.5.5 |
- Harness commit
d240badd966933fa19eec0606b2d1f3945f18ad2- Mock SHA-256
99d01f709d8d9c1f49055aa265d5eb7475264d6b513911fe317aee74a71190f8- Load generator SHA-256
not recorded- net.core.somaxconn
- 65,535
- net.ipv4.tcp_max_syn_backlog
- 65,535
Stream sweep diagnostics
Per streaming-sweep concurrency, as recorded by the engine: lanes by final HTTP status, and lanes that received no response.
| Ingress → upstream | Concurrency | Lanes by HTTP status | Lanes with no response |
|---|---|---|---|
| OpenAI → Anthropic | 1 | HTTP 200: 1 | 0 |
| OpenAI → Anthropic | 2 | HTTP 200: 2 | 0 |
| OpenAI → Anthropic | 4 | HTTP 200: 4 | 0 |
| OpenAI → Anthropic | 8 | HTTP 200: 8 | 0 |
| OpenAI → Anthropic | 16 | HTTP 200: 16 | 0 |
| OpenAI → Anthropic | 32 | HTTP 200: 32 | 0 |
| OpenAI → Anthropic | 64 | HTTP 200: 64 | 0 |
| OpenAI → Anthropic | 128 | HTTP 200: 128 | 0 |
| OpenAI → Anthropic | 256 | HTTP 200: 256 | 0 |
| OpenAI → Anthropic | 512 | HTTP 200: 512 | 0 |
| OpenAI → Anthropic | 1,024 | HTTP 200: 1,024 | 0 |
| OpenAI → Anthropic | 2,048 | HTTP 200: 2,048 | 0 |
| OpenAI → Anthropic | 1,536 | HTTP 200: 1,536 | 0 |
| OpenAI → Anthropic | 1,792 | HTTP 200: 1,792 | 0 |
| OpenAI → Anthropic | 1,664 | HTTP 200: 1,664 | 0 |
| OpenAI → Anthropic | 1,600 | HTTP 200: 1,600 | 0 |
| OpenAI → Anthropic | 1,568 | HTTP 200: 1,568 | 0 |
| OpenAI → Anthropic | 1,552 | HTTP 200: 1,552 | 0 |
| OpenAI → Anthropic | 1,544 | HTTP 200: 1,544 | 0 |
| OpenAI → Anthropic | 1,540 | HTTP 200: 1,540 | 0 |
| OpenAI → Anthropic | 1,538 | HTTP 200: 1,538 | 0 |
| OpenAI → Anthropic | 1,537 | HTTP 200: 1,537 | 0 |
| OpenAI → Anthropic | 1,536 | HTTP 200: 1,536 | 0 |
| OpenAI → Anthropic | 1,536 | HTTP 200: 1,536 | 0 |
| Anthropic → Anthropic | 1 | HTTP 200: 1 | 0 |
| Anthropic → Anthropic | 2 | HTTP 200: 2 | 0 |
| Anthropic → Anthropic | 4 | HTTP 200: 4 | 0 |
| Anthropic → Anthropic | 8 | HTTP 200: 8 | 0 |
| Anthropic → Anthropic | 16 | HTTP 200: 16 | 0 |
| Anthropic → Anthropic | 32 | HTTP 200: 32 | 0 |
| Anthropic → Anthropic | 64 | HTTP 200: 64 | 0 |
| Anthropic → Anthropic | 128 | HTTP 200: 128 | 0 |
| Anthropic → Anthropic | 256 | HTTP 200: 256 | 0 |
| Anthropic → Anthropic | 512 | HTTP 200: 512 | 0 |
| Anthropic → Anthropic | 1,024 | HTTP 200: 1,024 | 0 |
| Anthropic → Anthropic | 2,048 | HTTP 200: 2,048 | 0 |
| Anthropic → Anthropic | 1,536 | HTTP 200: 1,536 | 0 |
| Anthropic → Anthropic | 1,792 | HTTP 200: 1,792 | 0 |
| Anthropic → Anthropic | 1,664 | HTTP 200: 1,664 | 0 |
| Anthropic → Anthropic | 1,728 | HTTP 200: 1,728 | 0 |
| Anthropic → Anthropic | 1,696 | HTTP 200: 1,696 | 0 |
| Anthropic → Anthropic | 1,712 | HTTP 200: 1,712 | 0 |
| Anthropic → Anthropic | 1,704 | HTTP 200: 1,704 | 0 |
| Anthropic → Anthropic | 1,700 | HTTP 200: 1,700 | 0 |
| Anthropic → Anthropic | 1,698 | HTTP 200: 1,698 | 0 |
| Anthropic → Anthropic | 1,699 | HTTP 200: 1,699 | 0 |
| Anthropic → Anthropic | 1,698 | HTTP 200: 1,698 | 0 |
| Anthropic → Anthropic | 1,698 | HTTP 200: 1,698 | 0 |
| OpenAI → OpenAI | 1 | HTTP 200: 1 | 0 |
| OpenAI → OpenAI | 2 | HTTP 200: 2 | 0 |
| OpenAI → OpenAI | 4 | HTTP 200: 4 | 0 |
| OpenAI → OpenAI | 8 | HTTP 200: 8 | 0 |
| OpenAI → OpenAI | 16 | HTTP 200: 16 | 0 |
| OpenAI → OpenAI | 32 | HTTP 200: 32 | 0 |
| OpenAI → OpenAI | 64 | HTTP 200: 64 | 0 |
| OpenAI → OpenAI | 128 | HTTP 200: 128 | 0 |
| OpenAI → OpenAI | 256 | HTTP 200: 256 | 0 |
| OpenAI → OpenAI | 512 | HTTP 200: 512 | 0 |
| OpenAI → OpenAI | 1,024 | HTTP 200: 1,024 | 0 |
| OpenAI → OpenAI | 2,048 | HTTP 200: 2,048 | 0 |
| OpenAI → OpenAI | 1,536 | HTTP 200: 1,536 | 0 |
| OpenAI → OpenAI | 1,280 | HTTP 200: 1,280 | 0 |
| OpenAI → OpenAI | 1,408 | HTTP 200: 1,408 | 0 |
| OpenAI → OpenAI | 1,472 | HTTP 200: 1,472 | 0 |
| OpenAI → OpenAI | 1,440 | HTTP 200: 1,440 | 0 |
| OpenAI → OpenAI | 1,456 | HTTP 200: 1,456 | 0 |
| OpenAI → OpenAI | 1,448 | HTTP 200: 1,448 | 0 |
| OpenAI → OpenAI | 1,452 | HTTP 200: 1,452 | 0 |
| OpenAI → OpenAI | 1,454 | HTTP 200: 1,454 | 0 |
| OpenAI → OpenAI | 1,453 | HTTP 200: 1,453 | 0 |
| OpenAI → OpenAI | 1,452 | HTTP 200: 1,452 | 0 |
| OpenAI → OpenAI | 1,452 | HTTP 200: 1,452 | 0 |
| Anthropic → OpenAI | 1 | HTTP 200: 1 | 0 |
| Anthropic → OpenAI | 2 | HTTP 200: 2 | 0 |
| Anthropic → OpenAI | 4 | HTTP 200: 4 | 0 |
| Anthropic → OpenAI | 8 | HTTP 200: 8 | 0 |
| Anthropic → OpenAI | 16 | HTTP 200: 16 | 0 |
| Anthropic → OpenAI | 32 | HTTP 200: 32 | 0 |
| Anthropic → OpenAI | 64 | HTTP 200: 64 | 0 |
| Anthropic → OpenAI | 128 | HTTP 200: 128 | 0 |
| Anthropic → OpenAI | 256 | HTTP 200: 256 | 0 |
| Anthropic → OpenAI | 512 | HTTP 200: 512 | 0 |
| Anthropic → OpenAI | 1,024 | HTTP 200: 1,024 | 0 |
| Anthropic → OpenAI | 2,048 | HTTP 200: 2,048 | 0 |
| Anthropic → OpenAI | 1,536 | HTTP 200: 1,536 | 0 |
| Anthropic → OpenAI | 1,792 | HTTP 200: 1,792 | 0 |
| Anthropic → OpenAI | 1,664 | HTTP 200: 1,664 | 0 |
| Anthropic → OpenAI | 1,600 | HTTP 200: 1,600 | 0 |
| Anthropic → OpenAI | 1,632 | HTTP 200: 1,632 | 0 |
| Anthropic → OpenAI | 1,616 | HTTP 200: 1,616 | 0 |
| Anthropic → OpenAI | 1,608 | HTTP 200: 1,608 | 0 |
| Anthropic → OpenAI | 1,612 | HTTP 200: 1,612 | 0 |
| Anthropic → OpenAI | 1,610 | HTTP 200: 1,610 | 0 |
| Anthropic → OpenAI | 1,611 | HTTP 200: 1,611 | 0 |
| Anthropic → OpenAI | 1,610 | HTTP 200: 1,610 | 0 |
| Anthropic → OpenAI | 1,610 | HTTP 200: 1,610 | 0 |
Evidence for the current measurements
- run-busbar-16d2da7bc28375e55402b153f0f49b1550d5556f6ebb71eda60b8c3c47ad8d5f · complete raw record including RSS series
Configuration recorded by the run
# config.gen.yaml
listen: 127.0.0.1:8080
providers:
mock-openai:
api_key:
env: BENCH_MOCK_KEY
mock-responses:
api_key:
env: BENCH_MOCK_KEY
mock-anthropic:
api_key:
env: BENCH_MOCK_KEY
mock-gemini:
api_key:
env: BENCH_MOCK_KEY
mock-cohere:
api_key:
env: BENCH_MOCK_KEY
mock-bedrock:
api_key:
env: BENCH_BEDROCK_KEY
models:
gpt-4o-mini:
provider: mock-openai
gpt-4o-mini-responses:
provider: mock-responses
gpt-4o-mini-anthropic:
provider: mock-anthropic
gpt-4o-mini-gemini:
provider: mock-gemini
gpt-4o-mini-cohere:
provider: mock-cohere
gpt-4o-mini-bedrock:
provider: mock-bedrock
# providers.gen.yaml
mock-openai:
protocol: openai
base_url: http://127.0.0.1:8000
mock-responses:
protocol: responses
base_url: http://127.0.0.1:8000
mock-anthropic:
protocol: anthropic
base_url: http://127.0.0.1:8000
mock-gemini:
protocol: gemini
base_url: http://127.0.0.1:8000
mock-cohere:
protocol: cohere
base_url: http://127.0.0.1:8000
mock-bedrock:
protocol: bedrock
base_url: http://127.0.0.1:8000
Compare a shared protocol cell
- Busbar vs agentgateway
- Busbar vs AISIX (api7)
- Busbar vs APISIX
- Busbar vs Bifrost
- Busbar vs GoModel
- Busbar vs Helicone
- Busbar vs Higress
- Busbar vs Kong
- Busbar vs LiteLLM · Python
- Busbar vs LiteLLM · Rust
- Busbar vs One-API
- Busbar vs Plano
- Busbar vs Portkey
- Busbar vs TensorZero