AISIX (api7) 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: ghcr.io/api7/aisix:1.9.0. 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 → upstreamMetricValueState / reasonp99 bound (ms)
OpenAI → AnthropicAdded latency p50 (µs)162Valid · n=32,039 gateway / 261,797 direct—
OpenAI → AnthropicAdded latency p99 (µs)190Valid · n=32,039 gateway / 261,797 direct—
OpenAI → AnthropicGateway p99 at concurrency 1 (µs)218Valid · n=32,039—
OpenAI → AnthropicDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=261,797—
OpenAI → AnthropicCPU per request (µs)158.6Valid—
OpenAI → AnthropicRequests per CPU-second6,305.9Valid—
OpenAI → AnthropicCost window concurrency8Valid—
OpenAI → AnthropicCost window core utilisation0.786Valid—
OpenAI → AnthropicCost window successful requests119,308Valid—
OpenAI → AnthropicCost window throughput (req/s)19,882Valid—
OpenAI → AnthropicCost window threads16Valid—
OpenAI → AnthropicNon-voluntary context switches per request0.00000838Valid—
OpenAI → AnthropicCost window major page faults0Valid—
OpenAI → AnthropicAdded time to first event p50 (µs)294Valid · n=100 gateway / 100 direct—
OpenAI → AnthropicAdded time to first event p99 (µs)500Valid · n=100 gateway / 100 direct—
OpenAI → AnthropicAdded inter-frame gap p50 (µs)9Valid · n=63 gateway / 63 direct—
OpenAI → AnthropicAdded inter-frame gap p99 (µs)19Valid · n=63 gateway / 63 direct—
OpenAI → AnthropicStreams sustained (frames/s)34,767.8Valid—
OpenAI → AnthropicStreams sustained775Valid—
OpenAI → AnthropicIdle RSS (MiB)34.36Valid—
OpenAI → AnthropicSteady-state RSS (MiB)67.63Valid—
OpenAI → AnthropicRecovered RSS (MiB)54.02Valid—
OpenAI → AnthropicPeak RSS (MiB)69.21Valid—
OpenAI → AnthropicPeak RSS high-water mark (MiB)68.88Valid—
OpenAI → AnthropicTime to plateau (s)—Not measured—
OpenAI → AnthropicMemory growth (MiB/min)0.108Valid—
OpenAI → AnthropicMemory load duration (s)300Valid—
OpenAI → AnthropicMemory shape under load0Valid—
OpenAI → AnthropicIdle memory shape—Not measured—
OpenAI → AnthropicIdle memory static1Valid—
OpenAI → AnthropicIdle memory growth (MiB/min)-0.687Valid—
OpenAI → AnthropicThroughput (req/s)26,114concurrency 16; n=156,7051 ms
OpenAI → AnthropicThroughput (req/s)28,484concurrency 64; n=170,9155 ms
OpenAI → AnthropicThroughput (req/s)29,206concurrency 128; n=175,26610 ms
OpenAI → AnthropicThroughput (req/s)29,206concurrency 128; n=175,26650 ms
OpenAI → AnthropicThroughput (req/s)29,206concurrency 128; n=175,266100 ms
OpenAI → AnthropicThroughput (req/s)29,206concurrency 128; n=175,266Unbounded
Bedrock → Anthropic——Not declared: recorded note 1—
Gemini → Anthropic——Not declared: recorded note 1—
Anthropic → AnthropicAdded latency p50 (µs)150Valid · n=34,173 gateway / 257,073 direct—
Anthropic → AnthropicAdded latency p99 (µs)175Valid · n=34,173 gateway / 257,073 direct—
Anthropic → AnthropicGateway p99 at concurrency 1 (µs)204Valid · n=34,173—
Anthropic → AnthropicDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=257,073—
Anthropic → AnthropicCPU per request (µs)150.1Valid—
Anthropic → AnthropicRequests per CPU-second6,664.2Valid—
Anthropic → AnthropicCost window concurrency8Valid—
Anthropic → AnthropicCost window core utilisation0.993Valid—
Anthropic → AnthropicCost window successful requests159,340Valid—
Anthropic → AnthropicCost window throughput (req/s)26,553Valid—
Anthropic → AnthropicCost window threads16Valid—
Anthropic → AnthropicNon-voluntary context switches per request0Valid—
Anthropic → AnthropicCost window major page faults0Valid—
Anthropic → AnthropicAdded time to first event p50 (µs)271Valid · n=100 gateway / 100 direct—
Anthropic → AnthropicAdded time to first event p99 (µs)444Valid · n=100 gateway / 100 direct—
Anthropic → AnthropicAdded inter-frame gap p50 (µs)3Valid · n=63 gateway / 63 direct—
Anthropic → AnthropicAdded inter-frame gap p99 (µs)—Below resolution · n=63 gateway / 63 direct—
Anthropic → AnthropicStreams sustained (frames/s)32,295.2Valid—
Anthropic → AnthropicStreams sustained677Valid—
Anthropic → AnthropicIdle RSS (MiB)34.68Valid—
Anthropic → AnthropicSteady-state RSS (MiB)68.68Valid—
Anthropic → AnthropicRecovered RSS (MiB)52.76Valid—
Anthropic → AnthropicPeak RSS (MiB)70.72Valid—
Anthropic → AnthropicPeak RSS high-water mark (MiB)70.5Valid—
Anthropic → AnthropicTime to plateau (s)—Not measured—
Anthropic → AnthropicMemory growth (MiB/min)0.325Valid—
Anthropic → AnthropicMemory load duration (s)300Valid—
Anthropic → AnthropicMemory shape under load0Valid—
Anthropic → AnthropicIdle memory shape—Not measured—
Anthropic → AnthropicIdle memory static1Valid—
Anthropic → AnthropicIdle memory growth (MiB/min)-0.687Valid—
Anthropic → AnthropicThroughput (req/s)28,154concurrency 16; n=168,9461 ms
Anthropic → AnthropicThroughput (req/s)30,535concurrency 64; n=183,2205 ms
Anthropic → AnthropicThroughput (req/s)31,272concurrency 128; n=187,66110 ms
Anthropic → AnthropicThroughput (req/s)31,272concurrency 128; n=187,66150 ms
Anthropic → AnthropicThroughput (req/s)31,272concurrency 128; n=187,661100 ms
Anthropic → AnthropicThroughput (req/s)31,272concurrency 128; n=187,661Unbounded
OpenAI Responses → AnthropicAdded latency p50 (µs)182Valid · n=28,838 gateway / 259,566 direct—
OpenAI Responses → AnthropicAdded latency p99 (µs)215Valid · n=28,838 gateway / 259,566 direct—
OpenAI Responses → AnthropicGateway p99 at concurrency 1 (µs)244Valid · n=28,838—
OpenAI Responses → AnthropicDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=259,566—
OpenAI Responses → AnthropicCPU per request (µs)181.1Valid—
OpenAI Responses → AnthropicRequests per CPU-second5,521.2Valid—
OpenAI Responses → AnthropicCost window concurrency8Valid—
OpenAI Responses → AnthropicCost window core utilisation0.986Valid—
OpenAI Responses → AnthropicCost window successful requests131,074Valid—
OpenAI Responses → AnthropicCost window throughput (req/s)21,842Valid—
OpenAI Responses → AnthropicCost window threads16Valid—
OpenAI Responses → AnthropicNon-voluntary context switches per request0Valid—
OpenAI Responses → AnthropicCost window major page faults0Valid—
OpenAI Responses → AnthropicAdded time to first event p50 (µs)—Untestable—
OpenAI Responses → AnthropicAdded time to first event p99 (µs)—Untestable—
OpenAI Responses → AnthropicAdded inter-frame gap p50 (µs)—Untestable—
OpenAI Responses → AnthropicAdded inter-frame gap p99 (µs)—Untestable—
OpenAI Responses → AnthropicStreams sustained (frames/s)—Untestable—
OpenAI Responses → AnthropicStreams sustained—Untestable—
OpenAI Responses → AnthropicIdle RSS (MiB)34.46Valid—
OpenAI Responses → AnthropicSteady-state RSS (MiB)65.39Valid—
OpenAI Responses → AnthropicRecovered RSS (MiB)52.29Valid—
OpenAI Responses → AnthropicPeak RSS (MiB)67.24Valid—
OpenAI Responses → AnthropicPeak RSS high-water mark (MiB)67.08Valid—
OpenAI Responses → AnthropicTime to plateau (s)—Not measured—
OpenAI Responses → AnthropicMemory growth (MiB/min)-0.113Valid—
OpenAI Responses → AnthropicMemory load duration (s)300Valid—
OpenAI Responses → AnthropicMemory shape under load0Valid—
OpenAI Responses → AnthropicIdle memory shape-1Valid—
OpenAI Responses → AnthropicIdle memory static0Valid—
OpenAI Responses → AnthropicIdle memory growth (MiB/min)-0.832Valid—
OpenAI Responses → AnthropicThroughput (req/s)22,218concurrency 8; n=133,3281 ms
OpenAI Responses → AnthropicThroughput (req/s)25,257concurrency 64; n=151,5485 ms
OpenAI Responses → AnthropicThroughput (req/s)25,869concurrency 128; n=155,23910 ms
OpenAI Responses → AnthropicThroughput (req/s)25,869concurrency 128; n=155,23950 ms
OpenAI Responses → AnthropicThroughput (req/s)25,869concurrency 128; n=155,239100 ms
OpenAI Responses → AnthropicThroughput (req/s)25,869concurrency 128; n=155,239Unbounded
Cohere → Anthropic——Not declared: recorded note 1—
Cohere → OpenAI——Not declared: recorded note 1—
Bedrock → OpenAI——Not declared: recorded note 1—
Anthropic → OpenAIAdded latency p50 (µs)152Valid · n=34,038 gateway / 258,158 direct—
Anthropic → OpenAIAdded latency p99 (µs)178Valid · n=34,038 gateway / 258,158 direct—
Anthropic → OpenAIGateway p99 at concurrency 1 (µs)207Valid · n=34,038—
Anthropic → OpenAIDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=258,158—
Anthropic → OpenAICPU per request (µs)152.1Valid—
Anthropic → OpenAIRequests per CPU-second6,574.8Valid—
Anthropic → OpenAICost window concurrency8Valid—
Anthropic → OpenAICost window core utilisation0.756Valid—
Anthropic → OpenAICost window successful requests119,727Valid—
Anthropic → OpenAICost window throughput (req/s)19,952Valid—
Anthropic → OpenAICost window threads16Valid—
Anthropic → OpenAINon-voluntary context switches per request0Valid—
Anthropic → OpenAICost window major page faults0Valid—
Anthropic → OpenAIAdded time to first event p50 (µs)275Valid · n=100 gateway / 100 direct—
Anthropic → OpenAIAdded time to first event p99 (µs)461Valid · n=100 gateway / 100 direct—
Anthropic → OpenAIAdded inter-frame gap p50 (µs)—Below resolution · n=63 gateway / 63 direct—
Anthropic → OpenAIAdded inter-frame gap p99 (µs)—Below resolution · n=63 gateway / 63 direct—
Anthropic → OpenAIStreams sustained (frames/s)31,294.4Valid—
Anthropic → OpenAIStreams sustained658Valid—
Anthropic → OpenAIIdle RSS (MiB)34.22Valid—
Anthropic → OpenAISteady-state RSS (MiB)67.3Valid—
Anthropic → OpenAIRecovered RSS (MiB)51.81Valid—
Anthropic → OpenAIPeak RSS (MiB)68.92Valid—
Anthropic → OpenAIPeak RSS high-water mark (MiB)68.56Valid—
Anthropic → OpenAITime to plateau (s)—Not measured—
Anthropic → OpenAIMemory growth (MiB/min)-0.507Valid—
Anthropic → OpenAIMemory load duration (s)300Valid—
Anthropic → OpenAIMemory shape under load0Valid—
Anthropic → OpenAIIdle memory shape-1Valid—
Anthropic → OpenAIIdle memory static0Valid—
Anthropic → OpenAIIdle memory growth (MiB/min)-0.92Valid—
Anthropic → OpenAIThroughput (req/s)27,958concurrency 16; n=167,7691 ms
Anthropic → OpenAIThroughput (req/s)30,196concurrency 64; n=181,1845 ms
Anthropic → OpenAIThroughput (req/s)31,041concurrency 128; n=186,27810 ms
Anthropic → OpenAIThroughput (req/s)31,041concurrency 128; n=186,27850 ms
Anthropic → OpenAIThroughput (req/s)31,041concurrency 128; n=186,278100 ms
Anthropic → OpenAIThroughput (req/s)31,041concurrency 128; n=186,278Unbounded
Gemini → OpenAI——Not declared: recorded note 1—
OpenAI Responses → OpenAIAdded latency p50 (µs)169Valid · n=30,866 gateway / 260,253 direct—
OpenAI Responses → OpenAIAdded latency p99 (µs)199Valid · n=30,866 gateway / 260,253 direct—
OpenAI Responses → OpenAIGateway p99 at concurrency 1 (µs)228Valid · n=30,866—
OpenAI Responses → OpenAIDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=260,253—
OpenAI Responses → OpenAICPU per request (µs)162.1Valid—
OpenAI Responses → OpenAIRequests per CPU-second6,168.3Valid—
OpenAI Responses → OpenAICost window concurrency8Valid—
OpenAI Responses → OpenAICost window core utilisation0.785Valid—
OpenAI Responses → OpenAICost window successful requests116,520Valid—
OpenAI Responses → OpenAICost window throughput (req/s)19,417Valid—
OpenAI Responses → OpenAICost window threads16Valid—
OpenAI Responses → OpenAINon-voluntary context switches per request0Valid—
OpenAI Responses → OpenAICost window major page faults0Valid—
OpenAI Responses → OpenAIAdded time to first event p50 (µs)—Untestable—
OpenAI Responses → OpenAIAdded time to first event p99 (µs)—Untestable—
OpenAI Responses → OpenAIAdded inter-frame gap p50 (µs)—Untestable—
OpenAI Responses → OpenAIAdded inter-frame gap p99 (µs)—Untestable—
OpenAI Responses → OpenAIStreams sustained (frames/s)—Untestable—
OpenAI Responses → OpenAIStreams sustained—Untestable—
OpenAI Responses → OpenAIIdle RSS (MiB)34.47Valid—
OpenAI Responses → OpenAISteady-state RSS (MiB)66.86Valid—
OpenAI Responses → OpenAIRecovered RSS (MiB)52.7Valid—
OpenAI Responses → OpenAIPeak RSS (MiB)68.23Valid—
OpenAI Responses → OpenAIPeak RSS high-water mark (MiB)68.01Valid—
OpenAI Responses → OpenAITime to plateau (s)—Not measured—
OpenAI Responses → OpenAIMemory growth (MiB/min)1.24Valid—
OpenAI Responses → OpenAIMemory load duration (s)300Valid—
OpenAI Responses → OpenAIMemory shape under load0Valid—
OpenAI Responses → OpenAIIdle memory shape-1Valid—
OpenAI Responses → OpenAIIdle memory static0Valid—
OpenAI Responses → OpenAIIdle memory growth (MiB/min)-0.817Valid—
OpenAI Responses → OpenAIThroughput (req/s)25,449concurrency 16; n=152,7121 ms
OpenAI Responses → OpenAIThroughput (req/s)27,472concurrency 64; n=164,8355 ms
OpenAI Responses → OpenAIThroughput (req/s)28,458concurrency 128; n=170,77510 ms
OpenAI Responses → OpenAIThroughput (req/s)28,458concurrency 128; n=170,77550 ms
OpenAI Responses → OpenAIThroughput (req/s)28,458concurrency 128; n=170,775100 ms
OpenAI Responses → OpenAIThroughput (req/s)28,458concurrency 128; n=170,775Unbounded
OpenAI → OpenAIAdded latency p50 (µs)146Valid · n=35,048 gateway / 262,627 direct—
OpenAI → OpenAIAdded latency p99 (µs)174Valid · n=35,048 gateway / 262,627 direct—
OpenAI → OpenAIGateway p99 at concurrency 1 (µs)202Valid · n=35,048—
OpenAI → OpenAIDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=262,627—
OpenAI → OpenAICPU per request (µs)152.6Valid—
OpenAI → OpenAIRequests per CPU-second6,551.3Valid—
OpenAI → OpenAICost window concurrency8Valid—
OpenAI → OpenAICost window core utilisation0.945Valid—
OpenAI → OpenAICost window successful requests148,977Valid—
OpenAI → OpenAICost window throughput (req/s)24,826Valid—
OpenAI → OpenAICost window threads16Valid—
OpenAI → OpenAINon-voluntary context switches per request0Valid—
OpenAI → OpenAICost window major page faults0Valid—
OpenAI → OpenAIAdded time to first event p50 (µs)277Valid · n=100 gateway / 100 direct—
OpenAI → OpenAIAdded time to first event p99 (µs)366Valid · n=100 gateway / 100 direct—
OpenAI → OpenAIAdded inter-frame gap p50 (µs)2Valid · n=63 gateway / 63 direct—
OpenAI → OpenAIAdded inter-frame gap p99 (µs)—Below resolution · n=63 gateway / 63 direct—
OpenAI → OpenAIStreams sustained (frames/s)30,259.2Valid—
OpenAI → OpenAIStreams sustained645Valid—
OpenAI → OpenAIIdle RSS (MiB)34.29Valid—
OpenAI → OpenAISteady-state RSS (MiB)68.63Valid—
OpenAI → OpenAIRecovered RSS (MiB)52.3Valid—
OpenAI → OpenAIPeak RSS (MiB)70.43Valid—
OpenAI → OpenAIPeak RSS high-water mark (MiB)70.29Valid—
OpenAI → OpenAITime to plateau (s)—Not measured—
OpenAI → OpenAIMemory growth (MiB/min)0.376Valid—
OpenAI → OpenAIMemory load duration (s)300Valid—
OpenAI → OpenAIMemory shape under load0Valid—
OpenAI → OpenAIIdle memory shape-1Valid—
OpenAI → OpenAIIdle memory static0Valid—
OpenAI → OpenAIIdle memory growth (MiB/min)-0.853Valid—
OpenAI → OpenAIThroughput (req/s)28,888concurrency 16; n=173,3461 ms
OpenAI → OpenAIThroughput (req/s)31,374concurrency 64; n=188,2545 ms
OpenAI → OpenAIThroughput (req/s)32,092concurrency 128; n=192,58310 ms
OpenAI → OpenAIThroughput (req/s)32,092concurrency 128; n=192,58350 ms
OpenAI → OpenAIThroughput (req/s)32,092concurrency 128; n=192,583100 ms
OpenAI → OpenAIThroughput (req/s)32,092concurrency 128; n=192,583Unbounded
OpenAI Responses → Cohere——Not declared: recorded note 1—
OpenAI → Cohere——Not declared: recorded note 1—
Bedrock → Cohere——Not declared: recorded note 1—
Gemini → Cohere——Not declared: recorded note 1—
Cohere → Cohere——Not declared: recorded note 1—
Anthropic → Cohere——Not declared: recorded note 1—
Anthropic → GeminiAdded latency p50 (µs)162Valid · n=32,091 gateway / 257,892 direct—
Anthropic → GeminiAdded latency p99 (µs)187Valid · n=32,091 gateway / 257,892 direct—
Anthropic → GeminiGateway p99 at concurrency 1 (µs)216Valid · n=32,091—
Anthropic → GeminiDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=257,892—
Anthropic → GeminiCPU per request (µs)163.1Valid—
Anthropic → GeminiRequests per CPU-second6,130.4Valid—
Anthropic → GeminiCost window concurrency8Valid—
Anthropic → GeminiCost window core utilisation0.984Valid—
Anthropic → GeminiCost window successful requests145,168Valid—
Anthropic → GeminiCost window throughput (req/s)24,191Valid—
Anthropic → GeminiCost window threads16Valid—
Anthropic → GeminiNon-voluntary context switches per request0Valid—
Anthropic → GeminiCost window major page faults0Valid—
Anthropic → GeminiAdded time to first event p50 (µs)—Untestable—
Anthropic → GeminiAdded time to first event p99 (µs)—Untestable—
Anthropic → GeminiAdded inter-frame gap p50 (µs)—Untestable—
Anthropic → GeminiAdded inter-frame gap p99 (µs)—Untestable—
Anthropic → GeminiStreams sustained (frames/s)—Untestable—
Anthropic → GeminiStreams sustained—Untestable—
Anthropic → GeminiIdle RSS (MiB)34.72Valid—
Anthropic → GeminiSteady-state RSS (MiB)67.52Valid—
Anthropic → GeminiRecovered RSS (MiB)52.1Valid—
Anthropic → GeminiPeak RSS (MiB)69.29Valid—
Anthropic → GeminiPeak RSS high-water mark (MiB)69.11Valid—
Anthropic → GeminiTime to plateau (s)—Not measured—
Anthropic → GeminiMemory growth (MiB/min)0.615Valid—
Anthropic → GeminiMemory load duration (s)300Valid—
Anthropic → GeminiMemory shape under load0Valid—
Anthropic → GeminiIdle memory shape-1Valid—
Anthropic → GeminiIdle memory static0Valid—
Anthropic → GeminiIdle memory growth (MiB/min)-0.757Valid—
Anthropic → GeminiThroughput (req/s)26,504concurrency 16; n=159,0441 ms
Anthropic → GeminiThroughput (req/s)28,918concurrency 64; n=173,5145 ms
Anthropic → GeminiThroughput (req/s)29,709concurrency 128; n=178,28310 ms
Anthropic → GeminiThroughput (req/s)29,709concurrency 128; n=178,28350 ms
Anthropic → GeminiThroughput (req/s)29,709concurrency 128; n=178,283100 ms
Anthropic → GeminiThroughput (req/s)29,709concurrency 128; n=178,283Unbounded
Bedrock → Gemini——Not declared: recorded note 1—
Gemini → Gemini——Not declared: recorded note 1—
OpenAI → GeminiAdded latency p50 (µs)163Valid · n=31,889 gateway / 261,875 direct—
OpenAI → GeminiAdded latency p99 (µs)191Valid · n=31,889 gateway / 261,875 direct—
OpenAI → GeminiGateway p99 at concurrency 1 (µs)219Valid · n=31,889—
OpenAI → GeminiDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=261,875—
OpenAI → GeminiCPU per request (µs)160Valid—
OpenAI → GeminiRequests per CPU-second6,250.6Valid—
OpenAI → GeminiCost window concurrency8Valid—
OpenAI → GeminiCost window core utilisation0.982Valid—
OpenAI → GeminiCost window successful requests147,826Valid—
OpenAI → GeminiCost window throughput (req/s)24,634Valid—
OpenAI → GeminiCost window threads16Valid—
OpenAI → GeminiNon-voluntary context switches per request0Valid—
OpenAI → GeminiCost window major page faults0Valid—
OpenAI → GeminiAdded time to first event p50 (µs)—Untestable—
OpenAI → GeminiAdded time to first event p99 (µs)—Untestable—
OpenAI → GeminiAdded inter-frame gap p50 (µs)—Untestable—
OpenAI → GeminiAdded inter-frame gap p99 (µs)—Untestable—
OpenAI → GeminiStreams sustained (frames/s)—Untestable—
OpenAI → GeminiStreams sustained—Untestable—
OpenAI → GeminiIdle RSS (MiB)34.73Valid—
OpenAI → GeminiSteady-state RSS (MiB)67.99Valid—
OpenAI → GeminiRecovered RSS (MiB)53.27Valid—
OpenAI → GeminiPeak RSS (MiB)69.95Valid—
OpenAI → GeminiPeak RSS high-water mark (MiB)69.71Valid—
OpenAI → GeminiTime to plateau (s)—Not measured—
OpenAI → GeminiMemory growth (MiB/min)2.06Valid—
OpenAI → GeminiMemory load duration (s)300Valid—
OpenAI → GeminiMemory shape under load1Valid—
OpenAI → GeminiIdle memory shape—Not measured—
OpenAI → GeminiIdle memory static1Valid—
OpenAI → GeminiIdle memory growth (MiB/min)-0.425Valid—
OpenAI → GeminiThroughput (req/s)24,347concurrency 8; n=146,1001 ms
OpenAI → GeminiThroughput (req/s)29,193concurrency 64; n=175,1895 ms
OpenAI → GeminiThroughput (req/s)29,862concurrency 128; n=179,19810 ms
OpenAI → GeminiThroughput (req/s)29,862concurrency 128; n=179,19850 ms
OpenAI → GeminiThroughput (req/s)29,862concurrency 128; n=179,198100 ms
OpenAI → GeminiThroughput (req/s)29,862concurrency 128; n=179,198Unbounded
OpenAI Responses → GeminiAdded latency p50 (µs)179Valid · n=29,315 gateway / 264,153 direct—
OpenAI Responses → GeminiAdded latency p99 (µs)209Valid · n=29,315 gateway / 264,153 direct—
OpenAI Responses → GeminiGateway p99 at concurrency 1 (µs)237Valid · n=29,315—
OpenAI Responses → GeminiDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=264,153—
OpenAI Responses → GeminiCPU per request (µs)173.7Valid—
OpenAI Responses → GeminiRequests per CPU-second5,755.8Valid—
OpenAI Responses → GeminiCost window concurrency8Valid—
OpenAI Responses → GeminiCost window core utilisation0.984Valid—
OpenAI Responses → GeminiCost window successful requests136,412Valid—
OpenAI Responses → GeminiCost window throughput (req/s)22,732Valid—
OpenAI Responses → GeminiCost window threads16Valid—
OpenAI Responses → GeminiNon-voluntary context switches per request0Valid—
OpenAI Responses → GeminiCost window major page faults0Valid—
OpenAI Responses → GeminiAdded time to first event p50 (µs)—Untestable—
OpenAI Responses → GeminiAdded time to first event p99 (µs)—Untestable—
OpenAI Responses → GeminiAdded inter-frame gap p50 (µs)—Untestable—
OpenAI Responses → GeminiAdded inter-frame gap p99 (µs)—Untestable—
OpenAI Responses → GeminiStreams sustained (frames/s)—Untestable—
OpenAI Responses → GeminiStreams sustained—Untestable—
OpenAI Responses → GeminiIdle RSS (MiB)34.34Valid—
OpenAI Responses → GeminiSteady-state RSS (MiB)67.02Valid—
OpenAI Responses → GeminiRecovered RSS (MiB)52.66Valid—
OpenAI Responses → GeminiPeak RSS (MiB)68.57Valid—
OpenAI Responses → GeminiPeak RSS high-water mark (MiB)68.23Valid—
OpenAI Responses → GeminiTime to plateau (s)—Not measured—
OpenAI Responses → GeminiMemory growth (MiB/min)-0.081Valid—
OpenAI Responses → GeminiMemory load duration (s)300Valid—
OpenAI Responses → GeminiMemory shape under load0Valid—
OpenAI Responses → GeminiIdle memory shape-1Valid—
OpenAI Responses → GeminiIdle memory static0Valid—
OpenAI Responses → GeminiIdle memory growth (MiB/min)-0.873Valid—
OpenAI Responses → GeminiThroughput (req/s)24,971concurrency 16; n=149,8431 ms
OpenAI Responses → GeminiThroughput (req/s)27,134concurrency 64; n=162,8105 ms
OpenAI Responses → GeminiThroughput (req/s)27,902concurrency 128; n=167,43610 ms
OpenAI Responses → GeminiThroughput (req/s)27,902concurrency 128; n=167,43650 ms
OpenAI Responses → GeminiThroughput (req/s)27,902concurrency 128; n=167,436100 ms
OpenAI Responses → GeminiThroughput (req/s)27,902concurrency 128; n=167,436Unbounded
Cohere → Gemini——Not declared: recorded note 1—
OpenAI Responses → BedrockAdded latency p50 (µs)408Valid · n=13,815 gateway / 263,299 direct—
OpenAI Responses → BedrockAdded latency p99 (µs)458Valid · n=13,815 gateway / 263,299 direct—
OpenAI Responses → BedrockGateway p99 at concurrency 1 (µs)486Valid · n=13,815—
OpenAI Responses → BedrockDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=263,299—
OpenAI Responses → BedrockCPU per request (µs)417.1Valid—
OpenAI Responses → BedrockRequests per CPU-second2,397.4Valid—
OpenAI Responses → BedrockCost window concurrency8Valid—
OpenAI Responses → BedrockCost window core utilisation0.516Valid—
OpenAI Responses → BedrockCost window successful requests29,752Valid—
OpenAI Responses → BedrockCost window throughput (req/s)4,958Valid—
OpenAI Responses → BedrockCost window threads16Valid—
OpenAI Responses → BedrockNon-voluntary context switches per request0Valid—
OpenAI Responses → BedrockCost window major page faults0Valid—
OpenAI Responses → BedrockAdded time to first event p50 (µs)—Untestable—
OpenAI Responses → BedrockAdded time to first event p99 (µs)—Untestable—
OpenAI Responses → BedrockAdded inter-frame gap p50 (µs)—Untestable—
OpenAI Responses → BedrockAdded inter-frame gap p99 (µs)—Untestable—
OpenAI Responses → BedrockStreams sustained (frames/s)—Untestable—
OpenAI Responses → BedrockStreams sustained—Untestable—
OpenAI Responses → BedrockIdle RSS (MiB)34.14Valid—
OpenAI Responses → BedrockSteady-state RSS (MiB)63.89Valid—
OpenAI Responses → BedrockRecovered RSS (MiB)57.61Valid—
OpenAI Responses → BedrockPeak RSS (MiB)66.15Valid—
OpenAI Responses → BedrockPeak RSS high-water mark (MiB)65.87Valid—
OpenAI Responses → BedrockTime to plateau (s)—Not measured—
OpenAI Responses → BedrockMemory growth (MiB/min)2.17Valid—
OpenAI Responses → BedrockMemory load duration (s)300Valid—
OpenAI Responses → BedrockMemory shape under load1Valid—
OpenAI Responses → BedrockIdle memory shape-1Valid—
OpenAI Responses → BedrockIdle memory static0Valid—
OpenAI Responses → BedrockIdle memory growth (MiB/min)-0.775Valid—
OpenAI Responses → BedrockThroughput (req/s)6,806concurrency 4; n=40,8461 ms
OpenAI Responses → BedrockThroughput (req/s)8,726concurrency 16; n=52,3625 ms
OpenAI Responses → BedrockThroughput (req/s)9,937concurrency 32; n=59,63110 ms
OpenAI Responses → BedrockThroughput (req/s)10,099concurrency 64; n=60,59850 ms
OpenAI Responses → BedrockThroughput (req/s)10,099concurrency 64; n=60,598100 ms
OpenAI Responses → BedrockThroughput (req/s)10,099concurrency 64; n=60,598Unbounded
Gemini → Bedrock——Not declared: recorded note 1—
Bedrock → Bedrock——Not declared: recorded note 1—
OpenAI → BedrockAdded latency p50 (µs)381Valid · n=14,735 gateway / 261,517 direct—
OpenAI → BedrockAdded latency p99 (µs)427Valid · n=14,735 gateway / 261,517 direct—
OpenAI → BedrockGateway p99 at concurrency 1 (µs)455Valid · n=14,735—
OpenAI → BedrockDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=261,517—
OpenAI → BedrockCPU per request (µs)387.2Valid—
OpenAI → BedrockRequests per CPU-second2,582.6Valid—
OpenAI → BedrockCost window concurrency8Valid—
OpenAI → BedrockCost window core utilisation0.589Valid—
OpenAI → BedrockCost window successful requests36,595Valid—
OpenAI → BedrockCost window throughput (req/s)6,098Valid—
OpenAI → BedrockCost window threads16Valid—
OpenAI → BedrockNon-voluntary context switches per request0Valid—
OpenAI → BedrockCost window major page faults0Valid—
OpenAI → BedrockAdded time to first event p50 (µs)—Untestable—
OpenAI → BedrockAdded time to first event p99 (µs)—Untestable—
OpenAI → BedrockAdded inter-frame gap p50 (µs)—Untestable—
OpenAI → BedrockAdded inter-frame gap p99 (µs)—Untestable—
OpenAI → BedrockStreams sustained (frames/s)—Untestable—
OpenAI → BedrockStreams sustained—Untestable—
OpenAI → BedrockIdle RSS (MiB)34.75Valid—
OpenAI → BedrockSteady-state RSS (MiB)63.51Valid—
OpenAI → BedrockRecovered RSS (MiB)56.62Valid—
OpenAI → BedrockPeak RSS (MiB)64.69Valid—
OpenAI → BedrockPeak RSS high-water mark (MiB)64.43Valid—
OpenAI → BedrockTime to plateau (s)—Not measured—
OpenAI → BedrockMemory growth (MiB/min)0.349Valid—
OpenAI → BedrockMemory load duration (s)300Valid—
OpenAI → BedrockMemory shape under load0Valid—
OpenAI → BedrockIdle memory shape—Not measured—
OpenAI → BedrockIdle memory static1Valid—
OpenAI → BedrockIdle memory growth (MiB/min)-0.498Valid—
OpenAI → BedrockThroughput (req/s)4,228concurrency 2; n=25,3761 ms
OpenAI → BedrockThroughput (req/s)10,244concurrency 16; n=61,4745 ms
OpenAI → BedrockThroughput (req/s)10,358concurrency 32; n=62,16010 ms
OpenAI → BedrockThroughput (req/s)10,529concurrency 64; n=63,18250 ms
OpenAI → BedrockThroughput (req/s)10,529concurrency 64; n=63,182100 ms
OpenAI → BedrockThroughput (req/s)10,529concurrency 64; n=63,182Unbounded
Cohere → Bedrock——Not declared: recorded note 1—
Anthropic → BedrockAdded latency p50 (µs)400Valid · n=14,087 gateway / 261,607 direct—
Anthropic → BedrockAdded latency p99 (µs)448Valid · n=14,087 gateway / 261,607 direct—
Anthropic → BedrockGateway p99 at concurrency 1 (µs)476Valid · n=14,087—
Anthropic → BedrockDirect-to-mock p99 at concurrency 1 (µs)28Valid · n=261,607—
Anthropic → BedrockCPU per request (µs)389.4Valid—
Anthropic → BedrockRequests per CPU-second2,568Valid—
Anthropic → BedrockCost window concurrency8Valid—
Anthropic → BedrockCost window core utilisation0.514Valid—
Anthropic → BedrockCost window successful requests31,766Valid—
Anthropic → BedrockCost window throughput (req/s)5,293Valid—
Anthropic → BedrockCost window threads16Valid—
Anthropic → BedrockNon-voluntary context switches per request0Valid—
Anthropic → BedrockCost window major page faults0Valid—
Anthropic → BedrockAdded time to first event p50 (µs)—Untestable—
Anthropic → BedrockAdded time to first event p99 (µs)—Untestable—
Anthropic → BedrockAdded inter-frame gap p50 (µs)—Untestable—
Anthropic → BedrockAdded inter-frame gap p99 (µs)—Untestable—
Anthropic → BedrockStreams sustained (frames/s)—Untestable—
Anthropic → BedrockStreams sustained—Untestable—
Anthropic → BedrockIdle RSS (MiB)34.21Valid—
Anthropic → BedrockSteady-state RSS (MiB)62.88Valid—
Anthropic → BedrockRecovered RSS (MiB)56.08Valid—
Anthropic → BedrockPeak RSS (MiB)64.55Valid—
Anthropic → BedrockPeak RSS high-water mark (MiB)64.27Valid—
Anthropic → BedrockTime to plateau (s)—Not measured—
Anthropic → BedrockMemory growth (MiB/min)1.52Valid—
Anthropic → BedrockMemory load duration (s)300Valid—
Anthropic → BedrockMemory shape under load1Valid—
Anthropic → BedrockIdle memory shape—Not measured—
Anthropic → BedrockIdle memory static1Valid—
Anthropic → BedrockIdle memory growth (MiB/min)-0.409Valid—
Anthropic → BedrockThroughput (req/s)5,623concurrency 4; n=33,7491 ms
Anthropic → BedrockThroughput (req/s)10,290concurrency 32; n=61,7505 ms
Anthropic → BedrockThroughput (req/s)10,340concurrency 64; n=62,05210 ms
Anthropic → BedrockThroughput (req/s)10,340concurrency 64; n=62,05250 ms
Anthropic → BedrockThroughput (req/s)10,340concurrency 64; n=62,052100 ms
Anthropic → BedrockThroughput (req/s)10,340concurrency 64; n=62,052Unbounded
Gemini → OpenAI Responses——Not declared: recorded note 1—
Cohere → OpenAI Responses——Not declared: recorded note 1—
Bedrock → OpenAI Responses——Not declared: recorded note 1—
Anthropic → OpenAI Responses——Not declared: recorded note 1—
OpenAI Responses → OpenAI ResponsesAdded latency p50 (µs)160Valid · n=32,575 gateway / 260,507 direct—
OpenAI Responses → OpenAI ResponsesAdded latency p99 (µs)188Valid · n=32,575 gateway / 260,507 direct—
OpenAI Responses → OpenAI ResponsesGateway p99 at concurrency 1 (µs)217Valid · n=32,575—
OpenAI Responses → OpenAI ResponsesDirect-to-mock p99 at concurrency 1 (µs)29Valid · n=260,507—
OpenAI Responses → OpenAI ResponsesCPU per request (µs)148.1Valid—
OpenAI Responses → OpenAI ResponsesRequests per CPU-second6,752.7Valid—
OpenAI Responses → OpenAI ResponsesCost window concurrency8Valid—
OpenAI Responses → OpenAI ResponsesCost window core utilisation0.788Valid—
OpenAI Responses → OpenAI ResponsesCost window successful requests128,099Valid—
OpenAI Responses → OpenAI ResponsesCost window throughput (req/s)21,347Valid—
OpenAI Responses → OpenAI ResponsesCost window threads16Valid—
OpenAI Responses → OpenAI ResponsesNon-voluntary context switches per request0Valid—
OpenAI Responses → OpenAI ResponsesCost window major page faults0Valid—
OpenAI Responses → OpenAI ResponsesAdded time to first event p50 (µs)—Untestable—
OpenAI Responses → OpenAI ResponsesAdded time to first event p99 (µs)—Untestable—
OpenAI Responses → OpenAI ResponsesAdded inter-frame gap p50 (µs)—Untestable—
OpenAI Responses → OpenAI ResponsesAdded inter-frame gap p99 (µs)—Untestable—
OpenAI Responses → OpenAI ResponsesStreams sustained (frames/s)—Untestable—
OpenAI Responses → OpenAI ResponsesStreams sustained—Untestable—
OpenAI Responses → OpenAI ResponsesIdle RSS (MiB)34.33Valid—
OpenAI Responses → OpenAI ResponsesSteady-state RSS (MiB)67.35Valid—
OpenAI Responses → OpenAI ResponsesRecovered RSS (MiB)51.9Valid—
OpenAI Responses → OpenAI ResponsesPeak RSS (MiB)69.07Valid—
OpenAI Responses → OpenAI ResponsesPeak RSS high-water mark (MiB)68.7Valid—
OpenAI Responses → OpenAI ResponsesTime to plateau (s)—Not measured—
OpenAI Responses → OpenAI ResponsesMemory growth (MiB/min)-0.651Valid—
OpenAI Responses → OpenAI ResponsesMemory load duration (s)300Valid—
OpenAI Responses → OpenAI ResponsesMemory shape under load0Valid—
OpenAI Responses → OpenAI ResponsesIdle memory shape-1Valid—
OpenAI Responses → OpenAI ResponsesIdle memory static0Valid—
OpenAI Responses → OpenAI ResponsesIdle memory growth (MiB/min)-0.804Valid—
OpenAI Responses → OpenAI ResponsesThroughput (req/s)25,391concurrency 8; n=152,3681 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)30,114concurrency 64; n=180,6995 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)31,044concurrency 128; n=186,27010 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)31,044concurrency 128; n=186,27050 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)31,044concurrency 128; n=186,270100 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)31,044concurrency 128; n=186,270Unbounded
OpenAI → OpenAI Responses——Not declared: recorded note 1—

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 → upstreamCell stateProbe HTTP statusRecorded reasonRecorded detailEvidence scopeResponse excerptTested build
OpenAI → AnthropicServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Bedrock → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Gemini → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → AnthropicServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → AnthropicServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Cohere → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Cohere → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Bedrock → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → OpenAIServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Gemini → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → OpenAIServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
OpenAI → OpenAIServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Bedrock → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Gemini → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Cohere → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → GeminiServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Bedrock → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Gemini → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI → GeminiServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → GeminiServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Cohere → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → BedrockServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Gemini → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Bedrock → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI → BedrockServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Cohere → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → BedrockServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
Gemini → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Cohere → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Bedrock → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
Anthropic → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
OpenAI Responses → OpenAI ResponsesServed———Recorded run evidence—ghcr.io/api7/aisix:1.9.0
OpenAI → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/api7/aisix:1.9.0
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 → upstreamConcurrencyLanes by HTTP statusLanes with no response
OpenAI → Anthropic1HTTP 200: 10
OpenAI → Anthropic2HTTP 200: 20
OpenAI → Anthropic4HTTP 200: 40
OpenAI → Anthropic8HTTP 200: 80
OpenAI → Anthropic16HTTP 200: 160
OpenAI → Anthropic32HTTP 200: 320
OpenAI → Anthropic64HTTP 200: 640
OpenAI → Anthropic128HTTP 200: 1280
OpenAI → Anthropic256HTTP 200: 2560
OpenAI → Anthropic512HTTP 200: 5120
OpenAI → Anthropic1,024HTTP 200: 1,0240
OpenAI → Anthropic768HTTP 200: 7680
OpenAI → Anthropic896HTTP 200: 8960
OpenAI → Anthropic832HTTP 200: 8320
OpenAI → Anthropic800HTTP 200: 8000
OpenAI → Anthropic784HTTP 200: 7840
OpenAI → Anthropic776HTTP 200: 7760
OpenAI → Anthropic772HTTP 200: 7720
OpenAI → Anthropic774HTTP 200: 7740
OpenAI → Anthropic775HTTP 200: 7750
OpenAI → Anthropic775HTTP 200: 7750
OpenAI → Anthropic775HTTP 200: 7750
Anthropic → Anthropic1HTTP 200: 10
Anthropic → Anthropic2HTTP 200: 20
Anthropic → Anthropic4HTTP 200: 40
Anthropic → Anthropic8HTTP 200: 80
Anthropic → Anthropic16HTTP 200: 160
Anthropic → Anthropic32HTTP 200: 320
Anthropic → Anthropic64HTTP 200: 640
Anthropic → Anthropic128HTTP 200: 1280
Anthropic → Anthropic256HTTP 200: 2560
Anthropic → Anthropic512HTTP 200: 5120
Anthropic → Anthropic1,024HTTP 200: 1,0240
Anthropic → Anthropic768HTTP 200: 7680
Anthropic → Anthropic896HTTP 200: 8960
Anthropic → Anthropic832HTTP 200: 8320
Anthropic → Anthropic864HTTP 200: 8640
Anthropic → Anthropic848HTTP 200: 8480
Anthropic → Anthropic840HTTP 200: 8400
Anthropic → Anthropic844HTTP 200: 8440
Anthropic → Anthropic842HTTP 200: 8420
Anthropic → Anthropic843HTTP 200: 8430
Anthropic → Anthropic843HTTP 200: 8430
Anthropic → Anthropic843HTTP 200: 8430
Anthropic → Anthropic677HTTP 200: 6770
Anthropic → Anthropic677HTTP 200: 6770
Anthropic → Anthropic677HTTP 200: 6770
Anthropic → OpenAI1HTTP 200: 10
Anthropic → OpenAI2HTTP 200: 20
Anthropic → OpenAI4HTTP 200: 40
Anthropic → OpenAI8HTTP 200: 80
Anthropic → OpenAI16HTTP 200: 160
Anthropic → OpenAI32HTTP 200: 320
Anthropic → OpenAI64HTTP 200: 640
Anthropic → OpenAI128HTTP 200: 1280
Anthropic → OpenAI256HTTP 200: 2560
Anthropic → OpenAI512HTTP 200: 5120
Anthropic → OpenAI1,024HTTP 200: 1,0240
Anthropic → OpenAI768HTTP 200: 7680
Anthropic → OpenAI896HTTP 200: 8960
Anthropic → OpenAI832HTTP 200: 8320
Anthropic → OpenAI800HTTP 200: 8000
Anthropic → OpenAI816HTTP 200: 8160
Anthropic → OpenAI808HTTP 200: 8080
Anthropic → OpenAI804HTTP 200: 8040
Anthropic → OpenAI806HTTP 200: 8060
Anthropic → OpenAI805HTTP 200: 8050
Anthropic → OpenAI804HTTP 200: 8040
Anthropic → OpenAI804HTTP 200: 8040
Anthropic → OpenAI658HTTP 200: 6580
Anthropic → OpenAI658HTTP 200: 6580
Anthropic → OpenAI658HTTP 200: 6580
OpenAI → OpenAI1HTTP 200: 10
OpenAI → OpenAI2HTTP 200: 20
OpenAI → OpenAI4HTTP 200: 40
OpenAI → OpenAI8HTTP 200: 80
OpenAI → OpenAI16HTTP 200: 160
OpenAI → OpenAI32HTTP 200: 320
OpenAI → OpenAI64HTTP 200: 640
OpenAI → OpenAI128HTTP 200: 1280
OpenAI → OpenAI256HTTP 200: 2560
OpenAI → OpenAI512HTTP 200: 5120
OpenAI → OpenAI1,024HTTP 200: 1,0240
OpenAI → OpenAI768HTTP 200: 7680
OpenAI → OpenAI896HTTP 200: 8960
OpenAI → OpenAI832HTTP 200: 8320
OpenAI → OpenAI800HTTP 200: 8000
OpenAI → OpenAI784HTTP 200: 7840
OpenAI → OpenAI776HTTP 200: 7760
OpenAI → OpenAI780HTTP 200: 7800
OpenAI → OpenAI778HTTP 200: 7780
OpenAI → OpenAI779HTTP 200: 7790
OpenAI → OpenAI779HTTP 200: 7790
OpenAI → OpenAI779HTTP 200: 7790
OpenAI → OpenAI645HTTP 200: 6450
OpenAI → OpenAI645HTTP 200: 6450
OpenAI → OpenAI645HTTP 200: 6450

Recorded notes

Text recorded with the cells above. Each note is shown once; rows refer to it by number.

  1. Pinned AISIX v1.9.0 source 7e070f0bde3f653c3b197ed18c5b9205b866a4c5: README native adapter families, aisix-proxy/src/lib.rs routes and dispatch.rs::serves_natively_with, provider-vertex/bedrock bridge configuration. OpenAI, Responses and Anthropic ingress translate to Chat Completions, Anthropic, Gemini and Bedrock; only Responses ingress selects the Responses upstream. Separate provider keys use explicit apis:{} for Chat-only routing and the native Responses declaration default for the Responses alias. Thirteen exact pairs passed local recorder checks. Evidence: results/diagnostics/refresh-configuration/aisix-protocol-preflight.json
  2. Historical capability research, not a fresh probe of the tested build. Referenced versions may differ from gateway_build; do not infer current capability.

Evidence for the current measurements

Configuration recorded by the run

# config.gen.yaml
resources_file: "/etc/aisix/resources.yaml"
proxy:
  addr: "0.0.0.0:3000"
admin:
  admin_keys:
    - "aisix-admin-dummy"


# resources.gen.yaml
_format_version: "1"
provider_keys:
  - display_name: mock-openai
    provider: openai
    adapter: openai
    apis: {}
    api_base: "http://127.0.0.1:8000/v1"
    api_key: "sk-mock"
  - display_name: mock-anthropic
    provider: anthropic
    adapter: anthropic
    api_base: "http://127.0.0.1:8000"
    api_key: "sk-mock"
  - display_name: mock-vertex
    provider: google
    adapter: vertex
    api_base: "http://127.0.0.1:8000"
    api_key: '{"access_token":"dummy","project":"mock","region":"us-central1"}'
  - display_name: mock-bedrock
    provider: bedrock
    adapter: bedrock
    api_base: "http://127.0.0.1:8000"
    api_key: '{"access_key_id":"AKIAMOCKACCESSKEY","secret_access_key":"mock-secret","region":"us-east-1"}'
  - display_name: mock-responses
    provider: openai
    adapter: openai
    api_base: "http://127.0.0.1:8000/v1"
    api_key: "sk-mock"
models:
  - display_name: gpt-4o-mini
    provider: openai
    model_name: gpt-4o-mini
    provider_key: mock-openai
  - display_name: claude-3-5-sonnet-20241022
    provider: anthropic
    model_name: claude-3-5-sonnet-20241022
    provider_key: mock-anthropic
  - display_name: gemini-1.5-pro
    provider: google
    model_name: gemini-1.5-pro
    provider_key: mock-vertex
  - display_name: amazon.nova-micro-v1:0
    provider: bedrock
    model_name: amazon.nova-micro-v1:0
    provider_key: mock-bedrock
  - display_name: mock-responses
    provider: openai
    model_name: gpt-4o-mini
    provider_key: mock-responses
api_keys:
  - display_name: bench
    key_env: BENCH_AISIX_KEY
    allowed_models: ["*"]

Compare a shared protocol cell

Historical measurements