agentgateway 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/agentgateway/agentgateway:v1.7.0. 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 Responses → OpenAI ResponsesAdded latency p50 (µs)210Valid · n=25,353 gateway / 241,872 direct—
OpenAI Responses → OpenAI ResponsesAdded latency p99 (µs)247Valid · n=25,353 gateway / 241,872 direct—
OpenAI Responses → OpenAI ResponsesGateway p99 at concurrency 1 (µs)277Valid · n=25,353—
OpenAI Responses → OpenAI ResponsesDirect-to-mock p99 at concurrency 1 (µs)30Valid · n=241,872—
OpenAI Responses → OpenAI ResponsesCPU per request (µs)195.9Valid—
OpenAI Responses → OpenAI ResponsesRequests per CPU-second5,105.8Valid—
OpenAI Responses → OpenAI ResponsesCost window concurrency8Valid—
OpenAI Responses → OpenAI ResponsesCost window core utilisation0.982Valid—
OpenAI Responses → OpenAI ResponsesCost window successful requests120,751Valid—
OpenAI Responses → OpenAI ResponsesCost window throughput (req/s)20,122Valid—
OpenAI Responses → OpenAI ResponsesCost window threads13Valid—
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)34.46Valid—
OpenAI Responses → OpenAI ResponsesSteady-state RSS (MiB)53.24Valid—
OpenAI Responses → OpenAI ResponsesRecovered RSS (MiB)50.75Valid—
OpenAI Responses → OpenAI ResponsesPeak RSS (MiB)54.41Valid—
OpenAI Responses → OpenAI ResponsesPeak RSS high-water mark (MiB)54.24Valid—
OpenAI Responses → OpenAI ResponsesTime to plateau (s)—Not measured—
OpenAI Responses → OpenAI ResponsesMemory growth (MiB/min)0.764Valid—
OpenAI Responses → OpenAI ResponsesMemory load duration (s)300Valid—
OpenAI Responses → OpenAI ResponsesMemory shape under load0Valid—
OpenAI Responses → OpenAI ResponsesIdle memory shape—Not measured—
OpenAI Responses → OpenAI ResponsesIdle memory static1Valid—
OpenAI Responses → OpenAI ResponsesIdle memory growth (MiB/min)0.032Valid—
OpenAI Responses → OpenAI ResponsesThroughput (req/s)19,964concurrency 8; n=119,8031 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)20,925concurrency 64; n=125,5545 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)21,051concurrency 128; n=126,32710 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)21,224concurrency 256; n=127,35950 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)21,224concurrency 256; n=127,359100 ms
OpenAI Responses → OpenAI ResponsesThroughput (req/s)21,224concurrency 256; n=127,359Unbounded
Cohere → OpenAI Responses——Not declared: recorded note 1—
OpenAI → OpenAI Responses——Not declared: recorded note 1—
Anthropic → OpenAI ResponsesAdded latency p50 (µs)227Valid · n=23,610 gateway / 229,535 direct—
Anthropic → OpenAI ResponsesAdded latency p99 (µs)267Valid · n=23,610 gateway / 229,535 direct—
Anthropic → OpenAI ResponsesGateway p99 at concurrency 1 (µs)299Valid · n=23,610—
Anthropic → OpenAI ResponsesDirect-to-mock p99 at concurrency 1 (µs)32Valid · n=229,535—
Anthropic → OpenAI ResponsesCPU per request (µs)207.8Valid—
Anthropic → OpenAI ResponsesRequests per CPU-second4,813.2Valid—
Anthropic → OpenAI ResponsesCost window concurrency8Valid—
Anthropic → OpenAI ResponsesCost window core utilisation0.984Valid—
Anthropic → OpenAI ResponsesCost window successful requests114,025Valid—
Anthropic → OpenAI ResponsesCost window throughput (req/s)19,002Valid—
Anthropic → OpenAI ResponsesCost window threads13Valid—
Anthropic → OpenAI ResponsesNon-voluntary context switches per request0Valid—
Anthropic → OpenAI ResponsesCost window major page faults0Valid—
Anthropic → OpenAI ResponsesAdded time to first event p50 (µs)—Untestable—
Anthropic → OpenAI ResponsesAdded time to first event p99 (µs)—Untestable—
Anthropic → OpenAI ResponsesAdded inter-frame gap p50 (µs)—Untestable—
Anthropic → OpenAI ResponsesAdded inter-frame gap p99 (µs)—Untestable—
Anthropic → OpenAI ResponsesStreams sustained (frames/s)—Untestable—
Anthropic → OpenAI ResponsesStreams sustained—Untestable—
Anthropic → OpenAI ResponsesIdle RSS (MiB)34.34Valid—
Anthropic → OpenAI ResponsesSteady-state RSS (MiB)54.75Valid—
Anthropic → OpenAI ResponsesRecovered RSS (MiB)51.77Valid—
Anthropic → OpenAI ResponsesPeak RSS (MiB)55.42Valid—
Anthropic → OpenAI ResponsesPeak RSS high-water mark (MiB)55.16Valid—
Anthropic → OpenAI ResponsesTime to plateau (s)252Valid—
Anthropic → OpenAI ResponsesMemory growth (MiB/min)0.176Valid—
Anthropic → OpenAI ResponsesMemory load duration (s)300Valid—
Anthropic → OpenAI ResponsesMemory shape under load—Not measured—
Anthropic → OpenAI ResponsesIdle memory shape—Not measured—
Anthropic → OpenAI ResponsesIdle memory static1Valid—
Anthropic → OpenAI ResponsesIdle memory growth (MiB/min)0.049Valid—
Anthropic → OpenAI ResponsesThroughput (req/s)18,964concurrency 8; n=113,7981 ms
Anthropic → OpenAI ResponsesThroughput (req/s)19,692concurrency 32; n=118,1685 ms
Anthropic → OpenAI ResponsesThroughput (req/s)19,772concurrency 64; n=118,63510 ms
Anthropic → OpenAI ResponsesThroughput (req/s)20,024concurrency 256; n=120,16150 ms
Anthropic → OpenAI ResponsesThroughput (req/s)20,024concurrency 256; n=120,161100 ms
Anthropic → OpenAI ResponsesThroughput (req/s)20,024concurrency 256; n=120,161Unbounded
Bedrock → OpenAI Responses——Not declared: recorded note 1—
Gemini → OpenAI Responses——Not declared: recorded note 1—
Anthropic → BedrockAdded latency p50 (µs)218Valid · n=24,390 gateway / 237,655 direct—
Anthropic → BedrockAdded latency p99 (µs)256Valid · n=24,390 gateway / 237,655 direct—
Anthropic → BedrockGateway p99 at concurrency 1 (µs)287Valid · n=24,390—
Anthropic → BedrockDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=237,655—
Anthropic → BedrockCPU per request (µs)200.2Valid—
Anthropic → BedrockRequests per CPU-second4,994.8Valid—
Anthropic → BedrockCost window concurrency8Valid—
Anthropic → BedrockCost window core utilisation0.982Valid—
Anthropic → BedrockCost window successful requests118,127Valid—
Anthropic → BedrockCost window throughput (req/s)19,685Valid—
Anthropic → BedrockCost window threads13Valid—
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)34.07Valid—
Anthropic → BedrockSteady-state RSS (MiB)54.66Valid—
Anthropic → BedrockRecovered RSS (MiB)51.24Valid—
Anthropic → BedrockPeak RSS (MiB)55.45Valid—
Anthropic → BedrockPeak RSS high-water mark (MiB)55.26Valid—
Anthropic → BedrockTime to plateau (s)—Not measured—
Anthropic → BedrockMemory growth (MiB/min)-1.3Valid—
Anthropic → BedrockMemory load duration (s)300Valid—
Anthropic → BedrockMemory shape under load-1Valid—
Anthropic → BedrockIdle memory shape—Not measured—
Anthropic → BedrockIdle memory static1Valid—
Anthropic → BedrockIdle memory growth (MiB/min)0.049Valid—
Anthropic → BedrockThroughput (req/s)19,560concurrency 8; n=117,3761 ms
Anthropic → BedrockThroughput (req/s)20,404concurrency 64; n=122,4335 ms
Anthropic → BedrockThroughput (req/s)20,404concurrency 64; n=122,43310 ms
Anthropic → BedrockThroughput (req/s)20,660concurrency 256; n=123,97150 ms
Anthropic → BedrockThroughput (req/s)20,660concurrency 256; n=123,971100 ms
Anthropic → BedrockThroughput (req/s)20,660concurrency 256; n=123,971Unbounded
Gemini → Bedrock——Not declared: recorded note 1—
Cohere → Bedrock——Not declared: recorded note 1—
OpenAI Responses → BedrockAdded latency p50 (µs)231Valid · n=23,233 gateway / 241,925 direct—
OpenAI Responses → BedrockAdded latency p99 (µs)270Valid · n=23,233 gateway / 241,925 direct—
OpenAI Responses → BedrockGateway p99 at concurrency 1 (µs)301Valid · n=23,233—
OpenAI Responses → BedrockDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=241,925—
OpenAI Responses → BedrockCPU per request (µs)208.7Valid—
OpenAI Responses → BedrockRequests per CPU-second4,791.3Valid—
OpenAI Responses → BedrockCost window concurrency8Valid—
OpenAI Responses → BedrockCost window core utilisation0.984Valid—
OpenAI Responses → BedrockCost window successful requests113,505Valid—
OpenAI Responses → BedrockCost window throughput (req/s)18,915Valid—
OpenAI Responses → BedrockCost window threads13Valid—
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)34.1Valid—
OpenAI Responses → BedrockSteady-state RSS (MiB)54.75Valid—
OpenAI Responses → BedrockRecovered RSS (MiB)51.99Valid—
OpenAI Responses → BedrockPeak RSS (MiB)55.73Valid—
OpenAI Responses → BedrockPeak RSS high-water mark (MiB)55.29Valid—
OpenAI Responses → BedrockTime to plateau (s)—Not measured—
OpenAI Responses → BedrockMemory growth (MiB/min)0.023Valid—
OpenAI Responses → BedrockMemory load duration (s)300Valid—
OpenAI Responses → BedrockMemory shape under load0Valid—
OpenAI Responses → BedrockIdle memory shape—Not measured—
OpenAI Responses → BedrockIdle memory static1Valid—
OpenAI Responses → BedrockIdle memory growth (MiB/min)0.014Valid—
OpenAI Responses → BedrockThroughput (req/s)18,692concurrency 8; n=112,1681 ms
OpenAI Responses → BedrockThroughput (req/s)19,412concurrency 32; n=116,4825 ms
OpenAI Responses → BedrockThroughput (req/s)19,442concurrency 64; n=116,65810 ms
OpenAI Responses → BedrockThroughput (req/s)19,736concurrency 256; n=118,43050 ms
OpenAI Responses → BedrockThroughput (req/s)19,736concurrency 256; n=118,430100 ms
OpenAI Responses → BedrockThroughput (req/s)19,736concurrency 256; n=118,430Unbounded
Bedrock → Bedrock——Not declared: recorded note 1—
OpenAI → BedrockAdded latency p50 (µs)219Valid · n=24,338 gateway / 241,582 direct—
OpenAI → BedrockAdded latency p99 (µs)256Valid · n=24,338 gateway / 241,582 direct—
OpenAI → BedrockGateway p99 at concurrency 1 (µs)286Valid · n=24,338—
OpenAI → BedrockDirect-to-mock p99 at concurrency 1 (µs)30Valid · n=241,582—
OpenAI → BedrockCPU per request (µs)205.6Valid—
OpenAI → BedrockRequests per CPU-second4,863.4Valid—
OpenAI → BedrockCost window concurrency8Valid—
OpenAI → BedrockCost window core utilisation0.984Valid—
OpenAI → BedrockCost window successful requests115,215Valid—
OpenAI → BedrockCost window throughput (req/s)19,200Valid—
OpenAI → BedrockCost window threads13Valid—
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)34.2Valid—
OpenAI → BedrockSteady-state RSS (MiB)53.94Valid—
OpenAI → BedrockRecovered RSS (MiB)50.91Valid—
OpenAI → BedrockPeak RSS (MiB)54.72Valid—
OpenAI → BedrockPeak RSS high-water mark (MiB)54.31Valid—
OpenAI → BedrockTime to plateau (s)—Not measured—
OpenAI → BedrockMemory growth (MiB/min)-0.637Valid—
OpenAI → BedrockMemory load duration (s)300Valid—
OpenAI → BedrockMemory shape under load-1Valid—
OpenAI → BedrockIdle memory shape—Not measured—
OpenAI → BedrockIdle memory static1Valid—
OpenAI → BedrockIdle memory growth (MiB/min)0.045Valid—
OpenAI → BedrockThroughput (req/s)19,159concurrency 8; n=114,9711 ms
OpenAI → BedrockThroughput (req/s)19,912concurrency 32; n=119,4875 ms
OpenAI → BedrockThroughput (req/s)20,007concurrency 64; n=120,06310 ms
OpenAI → BedrockThroughput (req/s)20,265concurrency 256; n=121,60950 ms
OpenAI → BedrockThroughput (req/s)20,265concurrency 256; n=121,609100 ms
OpenAI → BedrockThroughput (req/s)20,265concurrency 256; n=121,609Unbounded
Anthropic → Cohere——Not declared: recorded note 1—
OpenAI Responses → Cohere——Not declared: recorded note 1—
Gemini → Cohere——Not declared: recorded note 1—
Cohere → Cohere——Not declared: recorded note 1—
OpenAI → Cohere——Not declared: recorded note 1—
Bedrock → Cohere——Not declared: recorded note 1—
OpenAI Responses → OpenAIAdded latency p50 (µs)229Valid · n=23,446 gateway / 240,137 direct—
OpenAI Responses → OpenAIAdded latency p99 (µs)267Valid · n=23,446 gateway / 240,137 direct—
OpenAI Responses → OpenAIGateway p99 at concurrency 1 (µs)298Valid · n=23,446—
OpenAI Responses → OpenAIDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=240,137—
OpenAI Responses → OpenAICPU per request (µs)207.5Valid—
OpenAI Responses → OpenAIRequests per CPU-second4,818.3Valid—
OpenAI Responses → OpenAICost window concurrency8Valid—
OpenAI Responses → OpenAICost window core utilisation0.984Valid—
OpenAI Responses → OpenAICost window successful requests114,146Valid—
OpenAI Responses → OpenAICost window throughput (req/s)19,022Valid—
OpenAI Responses → OpenAICost window threads13Valid—
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)34.04Valid—
OpenAI Responses → OpenAISteady-state RSS (MiB)53.94Valid—
OpenAI Responses → OpenAIRecovered RSS (MiB)51.01Valid—
OpenAI Responses → OpenAIPeak RSS (MiB)54.8Valid—
OpenAI Responses → OpenAIPeak RSS high-water mark (MiB)54.57Valid—
OpenAI Responses → OpenAITime to plateau (s)—Not measured—
OpenAI Responses → OpenAIMemory growth (MiB/min)1.19Valid—
OpenAI Responses → OpenAIMemory load duration (s)300Valid—
OpenAI Responses → OpenAIMemory shape under load1Valid—
OpenAI Responses → OpenAIIdle memory shape—Not measured—
OpenAI Responses → OpenAIIdle memory static1Valid—
OpenAI Responses → OpenAIIdle memory growth (MiB/min)0.045Valid—
OpenAI Responses → OpenAIThroughput (req/s)19,011concurrency 8; n=114,0801 ms
OpenAI Responses → OpenAIThroughput (req/s)19,731concurrency 32; n=118,3995 ms
OpenAI Responses → OpenAIThroughput (req/s)19,868concurrency 64; n=119,22010 ms
OpenAI Responses → OpenAIThroughput (req/s)20,063concurrency 256; n=120,39650 ms
OpenAI Responses → OpenAIThroughput (req/s)20,063concurrency 256; n=120,396100 ms
OpenAI Responses → OpenAIThroughput (req/s)20,063concurrency 256; n=120,396Unbounded
Cohere → OpenAI——Not declared: recorded note 1—
OpenAI → OpenAIAdded latency p50 (µs)216Valid · n=24,618 gateway / 233,879 direct—
OpenAI → OpenAIAdded latency p99 (µs)251Valid · n=24,618 gateway / 233,879 direct—
OpenAI → OpenAIGateway p99 at concurrency 1 (µs)282Valid · n=24,618—
OpenAI → OpenAIDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=233,879—
OpenAI → OpenAICPU per request (µs)188.6Valid—
OpenAI → OpenAIRequests per CPU-second5,302.9Valid—
OpenAI → OpenAICost window concurrency8Valid—
OpenAI → OpenAICost window core utilisation0.972Valid—
OpenAI → OpenAICost window successful requests124,088Valid—
OpenAI → OpenAICost window throughput (req/s)20,679Valid—
OpenAI → OpenAICost window threads13Valid—
OpenAI → OpenAINon-voluntary context switches per request0Valid—
OpenAI → OpenAICost window major page faults0Valid—
OpenAI → OpenAIAdded time to first event p50 (µs)327Valid · n=100 gateway / 100 direct—
OpenAI → OpenAIAdded time to first event p99 (µs)360Valid · n=100 gateway / 100 direct—
OpenAI → OpenAIAdded inter-frame gap p50 (µs)10Valid · n=63 gateway / 63 direct—
OpenAI → OpenAIAdded inter-frame gap p99 (µs)—Below resolution · n=63 gateway / 63 direct—
OpenAI → OpenAIStreams sustained (frames/s)32,614.1Valid—
OpenAI → OpenAIStreams sustained2,048Valid—
OpenAI → OpenAIIdle RSS (MiB)34Valid—
OpenAI → OpenAISteady-state RSS (MiB)53.49Valid—
OpenAI → OpenAIRecovered RSS (MiB)50Valid—
OpenAI → OpenAIPeak RSS (MiB)54.52Valid—
OpenAI → OpenAIPeak RSS high-water mark (MiB)54.23Valid—
OpenAI → OpenAITime to plateau (s)—Not measured—
OpenAI → OpenAIMemory growth (MiB/min)-2.07Valid—
OpenAI → OpenAIMemory load duration (s)300Valid—
OpenAI → OpenAIMemory shape under load-1Valid—
OpenAI → OpenAIIdle memory shape—Not measured—
OpenAI → OpenAIIdle memory static1Valid—
OpenAI → OpenAIIdle memory growth (MiB/min)0.036Valid—
OpenAI → OpenAIThroughput (req/s)21,156concurrency 8; n=126,9521 ms
OpenAI → OpenAIThroughput (req/s)22,239concurrency 64; n=133,4475 ms
OpenAI → OpenAIThroughput (req/s)22,380concurrency 128; n=134,30410 ms
OpenAI → OpenAIThroughput (req/s)22,545concurrency 256; n=135,28350 ms
OpenAI → OpenAIThroughput (req/s)22,545concurrency 256; n=135,283100 ms
OpenAI → OpenAIThroughput (req/s)22,545concurrency 256; n=135,283Unbounded
Anthropic → OpenAIAdded latency p50 (µs)212Valid · n=25,009 gateway / 233,282 direct—
Anthropic → OpenAIAdded latency p99 (µs)247Valid · n=25,009 gateway / 233,282 direct—
Anthropic → OpenAIGateway p99 at concurrency 1 (µs)279Valid · n=25,009—
Anthropic → OpenAIDirect-to-mock p99 at concurrency 1 (µs)32Valid · n=233,282—
Anthropic → OpenAICPU per request (µs)196.3Valid—
Anthropic → OpenAIRequests per CPU-second5,093Valid—
Anthropic → OpenAICost window concurrency8Valid—
Anthropic → OpenAICost window core utilisation0.968Valid—
Anthropic → OpenAICost window successful requests118,717Valid—
Anthropic → OpenAICost window throughput (req/s)19,784Valid—
Anthropic → OpenAICost window threads13Valid—
Anthropic → OpenAINon-voluntary context switches per request0Valid—
Anthropic → OpenAICost window major page faults0Valid—
Anthropic → OpenAIAdded time to first event p50 (µs)341Valid · n=100 gateway / 100 direct—
Anthropic → OpenAIAdded time to first event p99 (µs)395Valid · n=100 gateway / 100 direct—
Anthropic → OpenAIAdded inter-frame gap p50 (µs)3Valid · n=63 gateway / 63 direct—
Anthropic → OpenAIAdded inter-frame gap p99 (µs)17Valid · n=63 gateway / 63 direct—
Anthropic → OpenAIStreams sustained (frames/s)—Not measured—
Anthropic → OpenAIStreams sustained—Not measured—
Anthropic → OpenAIIdle RSS (MiB)33.7Valid—
Anthropic → OpenAISteady-state RSS (MiB)53.57Valid—
Anthropic → OpenAIRecovered RSS (MiB)49.9Valid—
Anthropic → OpenAIPeak RSS (MiB)54.21Valid—
Anthropic → OpenAIPeak RSS high-water mark (MiB)54.05Valid—
Anthropic → OpenAITime to plateau (s)156Valid—
Anthropic → OpenAIMemory growth (MiB/min)0.1Valid—
Anthropic → OpenAIMemory load duration (s)300Valid—
Anthropic → OpenAIMemory shape under load—Not measured—
Anthropic → OpenAIIdle memory shape—Not measured—
Anthropic → OpenAIIdle memory static1Valid—
Anthropic → OpenAIIdle memory growth (MiB/min)0.049Valid—
Anthropic → OpenAIThroughput (req/s)19,933concurrency 8; n=119,6161 ms
Anthropic → OpenAIThroughput (req/s)20,841concurrency 64; n=125,0535 ms
Anthropic → OpenAIThroughput (req/s)20,997concurrency 128; n=126,00110 ms
Anthropic → OpenAIThroughput (req/s)21,100concurrency 256; n=126,61850 ms
Anthropic → OpenAIThroughput (req/s)21,100concurrency 256; n=126,618100 ms
Anthropic → OpenAIThroughput (req/s)21,100concurrency 256; n=126,618Unbounded
Bedrock → OpenAI——Not declared: recorded note 1—
Gemini → OpenAI——Not declared: recorded note 1—
Anthropic → AnthropicAdded latency p50 (µs)206Valid · n=25,747 gateway / 233,859 direct—
Anthropic → AnthropicAdded latency p99 (µs)240Valid · n=25,747 gateway / 233,859 direct—
Anthropic → AnthropicGateway p99 at concurrency 1 (µs)271Valid · n=25,747—
Anthropic → AnthropicDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=233,859—
Anthropic → AnthropicCPU per request (µs)191.3Valid—
Anthropic → AnthropicRequests per CPU-second5,226.9Valid—
Anthropic → AnthropicCost window concurrency8Valid—
Anthropic → AnthropicCost window core utilisation0.963Valid—
Anthropic → AnthropicCost window successful requests121,211Valid—
Anthropic → AnthropicCost window throughput (req/s)20,199Valid—
Anthropic → AnthropicCost window threads13Valid—
Anthropic → AnthropicNon-voluntary context switches per request0Valid—
Anthropic → AnthropicCost window major page faults0Valid—
Anthropic → AnthropicAdded time to first event p50 (µs)323Valid · n=100 gateway / 100 direct—
Anthropic → AnthropicAdded time to first event p99 (µs)362Valid · 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)29,596.7Valid—
Anthropic → AnthropicStreams sustained4,104Valid—
Anthropic → AnthropicIdle RSS (MiB)34.16Valid—
Anthropic → AnthropicSteady-state RSS (MiB)53.17Valid—
Anthropic → AnthropicRecovered RSS (MiB)50.18Valid—
Anthropic → AnthropicPeak RSS (MiB)54.41Valid—
Anthropic → AnthropicPeak RSS high-water mark (MiB)54.25Valid—
Anthropic → AnthropicTime to plateau (s)—Not measured—
Anthropic → AnthropicMemory growth (MiB/min)-0.237Valid—
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)0.049Valid—
Anthropic → AnthropicThroughput (req/s)20,502concurrency 8; n=123,0331 ms
Anthropic → AnthropicThroughput (req/s)21,665concurrency 64; n=130,0005 ms
Anthropic → AnthropicThroughput (req/s)21,795concurrency 128; n=130,77510 ms
Anthropic → AnthropicThroughput (req/s)21,982concurrency 256; n=131,91350 ms
Anthropic → AnthropicThroughput (req/s)21,982concurrency 256; n=131,913100 ms
Anthropic → AnthropicThroughput (req/s)21,982concurrency 256; n=131,913Unbounded
Bedrock → Anthropic——Not declared: recorded note 1—
OpenAI Responses → Anthropic——Not declared: recorded note 1—
OpenAI → AnthropicAdded latency p50 (µs)228Valid · n=23,411 gateway / 237,919 direct—
OpenAI → AnthropicAdded latency p99 (µs)268Valid · n=23,411 gateway / 237,919 direct—
OpenAI → AnthropicGateway p99 at concurrency 1 (µs)299Valid · n=23,411—
OpenAI → AnthropicDirect-to-mock p99 at concurrency 1 (µs)31Valid · n=237,919—
OpenAI → AnthropicCPU per request (µs)207Valid—
OpenAI → AnthropicRequests per CPU-second4,831.1Valid—
OpenAI → AnthropicCost window concurrency8Valid—
OpenAI → AnthropicCost window core utilisation0.968Valid—
OpenAI → AnthropicCost window successful requests112,613Valid—
OpenAI → AnthropicCost window throughput (req/s)18,767Valid—
OpenAI → AnthropicCost window threads13Valid—
OpenAI → AnthropicNon-voluntary context switches per request0Valid—
OpenAI → AnthropicCost window major page faults0Valid—
OpenAI → AnthropicAdded time to first event p50 (µs)405Valid · n=100 gateway / 100 direct—
OpenAI → AnthropicAdded time to first event p99 (µs)446Valid · n=100 gateway / 100 direct—
OpenAI → AnthropicAdded inter-frame gap p50 (µs)6Valid · n=63 gateway / 63 direct—
OpenAI → AnthropicAdded inter-frame gap p99 (µs)22Valid · n=63 gateway / 63 direct—
OpenAI → AnthropicStreams sustained (frames/s)18,079.4Valid—
OpenAI → AnthropicStreams sustained2,559Valid—
OpenAI → AnthropicIdle RSS (MiB)34.2Valid—
OpenAI → AnthropicSteady-state RSS (MiB)53.82Valid—
OpenAI → AnthropicRecovered RSS (MiB)50.53Valid—
OpenAI → AnthropicPeak RSS (MiB)54.71Valid—
OpenAI → AnthropicPeak RSS high-water mark (MiB)54.35Valid—
OpenAI → AnthropicTime to plateau (s)—Not measured—
OpenAI → AnthropicMemory growth (MiB/min)-0.764Valid—
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)0.041Valid—
OpenAI → AnthropicThroughput (req/s)19,127concurrency 8; n=114,7831 ms
OpenAI → AnthropicThroughput (req/s)19,852concurrency 32; n=119,1225 ms
OpenAI → AnthropicThroughput (req/s)19,973concurrency 64; n=119,86410 ms
OpenAI → AnthropicThroughput (req/s)20,228concurrency 256; n=121,38250 ms
OpenAI → AnthropicThroughput (req/s)20,228concurrency 256; n=121,382100 ms
OpenAI → AnthropicThroughput (req/s)20,228concurrency 256; n=121,382Unbounded
Cohere → Anthropic——Not declared: recorded note 1—
Gemini → Anthropic——Not declared: recorded note 1—
Cohere → Gemini——Not declared: recorded note 1—
Gemini → GeminiAdded latency p50 (µs)211Valid · n=25,203 gateway / 241,373 direct—
Gemini → GeminiAdded latency p99 (µs)248Valid · n=25,203 gateway / 241,373 direct—
Gemini → GeminiGateway p99 at concurrency 1 (µs)278Valid · n=25,203—
Gemini → GeminiDirect-to-mock p99 at concurrency 1 (µs)30Valid · n=241,373—
Gemini → GeminiCPU per request (µs)196.9Valid—
Gemini → GeminiRequests per CPU-second5,079Valid—
Gemini → GeminiCost window concurrency8Valid—
Gemini → GeminiCost window core utilisation0.982Valid—
Gemini → GeminiCost window successful requests120,067Valid—
Gemini → GeminiCost window throughput (req/s)20,008Valid—
Gemini → GeminiCost window threads13Valid—
Gemini → GeminiNon-voluntary context switches per request0Valid—
Gemini → GeminiCost window major page faults0Valid—
Gemini → GeminiAdded time to first event p50 (µs)—Untestable—
Gemini → GeminiAdded time to first event p99 (µs)—Untestable—
Gemini → GeminiAdded inter-frame gap p50 (µs)—Untestable—
Gemini → GeminiAdded inter-frame gap p99 (µs)—Untestable—
Gemini → GeminiStreams sustained (frames/s)—Untestable—
Gemini → GeminiStreams sustained—Untestable—
Gemini → GeminiIdle RSS (MiB)33.96Valid—
Gemini → GeminiSteady-state RSS (MiB)54.24Valid—
Gemini → GeminiRecovered RSS (MiB)50.96Valid—
Gemini → GeminiPeak RSS (MiB)54.77Valid—
Gemini → GeminiPeak RSS high-water mark (MiB)54.52Valid—
Gemini → GeminiTime to plateau (s)—Not measured—
Gemini → GeminiMemory growth (MiB/min)-0.178Valid—
Gemini → GeminiMemory load duration (s)300Valid—
Gemini → GeminiMemory shape under load0Valid—
Gemini → GeminiIdle memory shape—Not measured—
Gemini → GeminiIdle memory static1Valid—
Gemini → GeminiIdle memory growth (MiB/min)0.041Valid—
Gemini → GeminiThroughput (req/s)19,915concurrency 8; n=119,5081 ms
Gemini → GeminiThroughput (req/s)20,912concurrency 64; n=125,4815 ms
Gemini → GeminiThroughput (req/s)20,976concurrency 128; n=125,87710 ms
Gemini → GeminiThroughput (req/s)21,168concurrency 256; n=127,02350 ms
Gemini → GeminiThroughput (req/s)21,168concurrency 256; n=127,023100 ms
Gemini → GeminiThroughput (req/s)21,168concurrency 256; n=127,023Unbounded
OpenAI Responses → Gemini——Not declared: recorded note 1—
OpenAI → GeminiAdded latency p50 (µs)217Valid · n=24,639 gateway / 244,550 direct—
OpenAI → GeminiAdded latency p99 (µs)253Valid · n=24,639 gateway / 244,550 direct—
OpenAI → GeminiGateway p99 at concurrency 1 (µs)283Valid · n=24,639—
OpenAI → GeminiDirect-to-mock p99 at concurrency 1 (µs)30Valid · n=244,550—
OpenAI → GeminiCPU per request (µs)202.1Valid—
OpenAI → GeminiRequests per CPU-second4,947.7Valid—
OpenAI → GeminiCost window concurrency8Valid—
OpenAI → GeminiCost window core utilisation0.983Valid—
OpenAI → GeminiCost window successful requests117,111Valid—
OpenAI → GeminiCost window throughput (req/s)19,516Valid—
OpenAI → GeminiCost window threads13Valid—
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)34.06Valid—
OpenAI → GeminiSteady-state RSS (MiB)53.59Valid—
OpenAI → GeminiRecovered RSS (MiB)50.47Valid—
OpenAI → GeminiPeak RSS (MiB)54.25Valid—
OpenAI → GeminiPeak RSS high-water mark (MiB)53.98Valid—
OpenAI → GeminiTime to plateau (s)—Not measured—
OpenAI → GeminiMemory growth (MiB/min)-0.361Valid—
OpenAI → GeminiMemory load duration (s)300Valid—
OpenAI → GeminiMemory shape under load0Valid—
OpenAI → GeminiIdle memory shape—Not measured—
OpenAI → GeminiIdle memory static1Valid—
OpenAI → GeminiIdle memory growth (MiB/min)0.002Valid—
OpenAI → GeminiThroughput (req/s)19,532concurrency 8; n=117,2071 ms
OpenAI → GeminiThroughput (req/s)20,455concurrency 64; n=122,7405 ms
OpenAI → GeminiThroughput (req/s)20,455concurrency 64; n=122,74010 ms
OpenAI → GeminiThroughput (req/s)20,723concurrency 256; n=124,35050 ms
OpenAI → GeminiThroughput (req/s)20,723concurrency 256; n=124,350100 ms
OpenAI → GeminiThroughput (req/s)20,723concurrency 256; n=124,350Unbounded
Bedrock → Gemini——Not declared: recorded note 1—
Anthropic → Gemini——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 Responses → OpenAI ResponsesServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → OpenAI ResponsesServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → OpenAI ResponsesNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → BedrockServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI Responses → BedrockServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → BedrockNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → BedrockServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI Responses → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → CohereNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI Responses → OpenAIServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → OpenAIServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → OpenAIServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → OpenAINot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → AnthropicServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI Responses → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → AnthropicServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → AnthropicNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Cohere → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Gemini → GeminiServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI Responses → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
OpenAI → GeminiServed———Recorded run evidence—ghcr.io/agentgateway/agentgateway:v1.7.0
Bedrock → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
Anthropic → GeminiNot declared—recorded note 1recorded note 1recorded note 2—ghcr.io/agentgateway/agentgateway:v1.7.0
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 → 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 → OpenAI2,048HTTP 200: 2,0480
OpenAI → OpenAI4,096HTTP 200: 4,0960
OpenAI → OpenAI3,072HTTP 200: 3,0720
OpenAI → OpenAI2,560HTTP 200: 2,5600
OpenAI → OpenAI2,304HTTP 200: 2,3040
OpenAI → OpenAI2,432HTTP 200: 2,4320
OpenAI → OpenAI2,496HTTP 200: 2,4960
OpenAI → OpenAI2,528HTTP 200: 2,5280
OpenAI → OpenAI2,544HTTP 200: 2,5440
OpenAI → OpenAI2,552HTTP 200: 2,5520
OpenAI → OpenAI2,556HTTP 200: 2,5560
OpenAI → OpenAI2,558HTTP 200: 2,5580
OpenAI → OpenAI2,557HTTP 200: 2,5570
OpenAI → OpenAI2,556HTTP 200: 2,5560
OpenAI → OpenAI2,556HTTP 200: 2,5560
OpenAI → OpenAI2,302HTTP 200: 2,3020
OpenAI → OpenAI2,175HTTP 200: 2,1750
OpenAI → OpenAI2,111HTTP 200: 2,1110
OpenAI → OpenAI2,079HTTP 200: 2,0790
OpenAI → OpenAI2,048HTTP 200: 2,0480
OpenAI → OpenAI2,048HTTP 200: 2,0480
OpenAI → OpenAI2,048HTTP 200: 2,0480
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 → OpenAI2,048HTTP 200: 2,0480
Anthropic → OpenAI4,096HTTP 200: 4,0960
Anthropic → OpenAI8,192HTTP 200: 8,1920
Anthropic → OpenAI6,144HTTP 200: 6,1440
Anthropic → OpenAI5,120HTTP 200: 5,1200
Anthropic → OpenAI4,608HTTP 200: 4,6080
Anthropic → OpenAI4,352HTTP 200: 4,3520
Anthropic → OpenAI4,224HTTP 200: 4,2240
Anthropic → OpenAI4,160HTTP 200: 4,1600
Anthropic → OpenAI4,128HTTP 200: 4,1280
Anthropic → OpenAI4,144HTTP 200: 4,1440
Anthropic → OpenAI4,136HTTP 200: 3,945191
Anthropic → OpenAI4,132HTTP 200: 3,959173
Anthropic → OpenAI4,130HTTP 200: 4,1300
Anthropic → OpenAI4,129HTTP 200: 4,1290
Anthropic → OpenAI4,128HTTP 200: 3,234894
Anthropic → OpenAI4,128HTTP 200: 4,05771
Anthropic → OpenAI4,112HTTP 200: 3,242870
Anthropic → OpenAI4,104HTTP 200: 2,9571,147
Anthropic → OpenAI4,100HTTP 200: 3,860240
Anthropic → OpenAI4,098HTTP 200: 2,7041,394
Anthropic → OpenAI4,096HTTP 200: 3,404692
Anthropic → OpenAI4,096HTTP 200: 3,208888
Anthropic → OpenAI4,096HTTP 200: 3,250846
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 → Anthropic2,048HTTP 200: 2,0480
Anthropic → Anthropic4,096HTTP 200: 4,0960
Anthropic → Anthropic8,192HTTP 200: 8,1920
Anthropic → Anthropic6,144HTTP 200: 6,1440
Anthropic → Anthropic5,120HTTP 200: 5,1200
Anthropic → Anthropic4,608HTTP 200: 4,6080
Anthropic → Anthropic4,352HTTP 200: 4,3520
Anthropic → Anthropic4,224HTTP 200: 4,2240
Anthropic → Anthropic4,160HTTP 200: 4,1600
Anthropic → Anthropic4,128HTTP 200: 4,1280
Anthropic → Anthropic4,112HTTP 200: 4,1120
Anthropic → Anthropic4,104HTTP 200: 4,1040
Anthropic → Anthropic4,108HTTP 200: 4,1080
Anthropic → Anthropic4,106HTTP 200: 4,1060
Anthropic → Anthropic4,105HTTP 200: 4,1050
Anthropic → Anthropic4,104HTTP 200: 4,1040
Anthropic → Anthropic4,104HTTP 200: 3,923181
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 → Anthropic2,048HTTP 200: 2,0480
OpenAI → Anthropic4,096HTTP 200: 4,0960
OpenAI → Anthropic3,072HTTP 200: 3,0720
OpenAI → Anthropic2,560HTTP 200: 2,5600
OpenAI → Anthropic2,304HTTP 200: 2,3040
OpenAI → Anthropic2,432HTTP 200: 2,4320
OpenAI → Anthropic2,496HTTP 200: 2,4960
OpenAI → Anthropic2,528HTTP 200: 2,5280
OpenAI → Anthropic2,544HTTP 200: 2,5440
OpenAI → Anthropic2,552HTTP 200: 2,5520
OpenAI → Anthropic2,556HTTP 200: 2,5560
OpenAI → Anthropic2,558HTTP 200: 2,5580
OpenAI → Anthropic2,559HTTP 200: 2,5590
OpenAI → Anthropic2,559HTTP 200: 2,5590
OpenAI → Anthropic2,559HTTP 200: 2,5590

Recorded notes

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

  1. Pinned v1.7.0 image and source 26f5aebe99a36a037fe54f8080a608aca5547e15: CHAT_TRANSLATIONS in crates/agentgateway/src/llm/mod.rs defines supported conversions. Separate custom providers constrain Chat versus Responses wire. Gemini native and Chat-to-Gemini are supported; Messages/Responses use Google OpenAI compatibility instead of native Gemini and are not counted as Gemini wire. Cohere preset similarly uses OpenAI compatibility, not Cohere native chat. No native Cohere or Bedrock ingress conversion is registered. Local recorder and response checks establish 12 pairs. Responses mock token-details correction is required for Messages-to-Responses. Evidence: results/diagnostics/refresh-configuration/agentgateway-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

llm:
  port: 8080
  models:
  - name: gpt-4o-mini
    provider: openAI
    params:
      model: gpt-4o-mini
      apiKey: "sk-dummy-openai"
      baseUrl: "http://127.0.0.1:8000/v1"
  - name: claude-3-5-sonnet-20241022
    provider: anthropic
    params:
      model: claude-3-5-sonnet-20241022
      apiKey: "sk-ant-dummy"
      baseUrl: "http://127.0.0.1:8000/v1"
  - name: anthropic.claude-3-sonnet-20240229-v1:0
    provider: bedrock
    params:
      model: anthropic.claude-3-sonnet-20240229-v1:0
      awsRegion: us-east-1
      baseUrl: "http://127.0.0.1:8000/bedrock"
  - name: gemini-1.5-pro
    provider: gemini
    params:
      model: gemini-1.5-pro
      apiKey: dummy
      baseUrl: http://127.0.0.1:8000
  - name: bench-chat
    provider:
      custom:
        formats:
        - type: completions
    params:
      model: gpt-4o-mini
      apiKey: dummy
      baseUrl: http://127.0.0.1:8000/v1
  - name: bench-responses
    provider:
      custom:
        formats:
        - type: responses
    params:
      model: gpt-4o-mini
      apiKey: dummy
      baseUrl: http://127.0.0.1:8000/v1

Compare a shared protocol cell

Historical measurements