# APISIX 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](https://onthebench.ai/gateways/method) .

Other · recorded build: apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38. Measured: 2026-11-10T17:39:33Z.

[JSON](https://onthebench.ai/api/v1/gateways/apisix)  ·  [Markdown](https://onthebench.ai/gateways/apisix.md)  ·  [CSV](https://onthebench.ai/gateways/data.csv?gateway=apisix&compact=1)  ·  [Project source](https://github.com/apache/apisix)  ·  [Interactive dashboard](https://onthebench.ai/gateways)

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) |
| --- | --- | --- | --- | --- |
| bedrock → openai | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → openai | added_latency_p50_us | 179 | valid; gateway_c1_samples=28247; direct_c1_samples=260315 | — |
| openai → openai | added_latency_p99_us | 340 | valid; gateway_c1_samples=28247; direct_c1_samples=260315 | — |
| openai → openai | gateway_c1_p99_us | 368 | valid; gateway_c1_samples=28247 | — |
| openai → openai | direct_c1_p99_us | 28 | valid; direct_c1_samples=260315 | — |
| openai → openai | cpu_us_per_request | 185.658855916589 | valid | — |
| openai → openai | rps_per_cpu_second | 5386.223000583772 | valid | — |
| openai → openai | cost_window_conc | 8 | valid | — |
| openai → openai | cost_core_utilisation | 0.7114382453326605 | valid | — |
| openai → openai | cost_window_ok | 92266 | valid | — |
| openai → openai | cost_window_rps | 15375 | valid | — |
| openai → openai | cost_threads | 12 | valid | — |
| openai → openai | cost_nonvol_ctxt_per_request | 0.00045520560119654044 | valid | — |
| openai → openai | cost_majflt | 0 | valid | — |
| openai → openai | added_ttft_p50_us | 10584 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| openai → openai | added_ttft_p99_us | 11228 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| openai → openai | added_gap_p50_us | below_resolution | below_resolution; gap_gw_samples=63; gap_direct_samples=63 | — |
| openai → openai | added_gap_p99_us | 8877 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| openai → openai | streams_sustained_fps | 63882.44005721134 | valid | — |
| openai → openai | streams_sustained | 1470 | valid | — |
| openai → openai | idle_rss_mib | 191.2109375 | valid | — |
| openai → openai | steady_state_rss_mib | 228.22265625 | valid | — |
| openai → openai | recovered_rss_mib | 228.3515625 | valid | — |
| openai → openai | peak_rss_mib | 228.44140625 | valid | — |
| openai → openai | peak_rss_hwm_mib | 233.2109375 | valid | — |
| openai → openai | time_to_plateau_s | 66 | valid | — |
| openai → openai | growth_rate_mib_per_min | 0.050196090921940856 | valid | — |
| openai → openai | load_s | 300 | valid | — |
| openai → openai | shape | not_measured | not_measured | — |
| openai → openai | idle_shape | 0 | valid | — |
| openai → openai | idle_static | 0 | valid | — |
| openai → openai | idle_growth_rate_mib_per_min | -0.46382647366702207 | valid | — |
| openai → openai | throughput_rps | 12187 req/s | concurrency 4; samples 73136 | 1 |
| openai → openai | throughput_rps | 23089 req/s | concurrency 32; samples 138550 | 5 |
| openai → openai | throughput_rps | 25474 req/s | concurrency 64; samples 152851 | 10 |
| openai → openai | throughput_rps | 26590 req/s | concurrency 256; samples 159560 | 50 |
| openai → openai | throughput_rps | 26590 req/s | concurrency 256; samples 159560 | 100 |
| openai → openai | throughput_rps | 26590 req/s | concurrency 256; samples 159560 | unbounded |
| anthropic → openai | added_latency_p50_us | 193 | valid; gateway_c1_samples=26470; direct_c1_samples=257076 | — |
| anthropic → openai | added_latency_p99_us | 373 | valid; gateway_c1_samples=26470; direct_c1_samples=257076 | — |
| anthropic → openai | gateway_c1_p99_us | 402 | valid; gateway_c1_samples=26470 | — |
| anthropic → openai | direct_c1_p99_us | 29 | valid; direct_c1_samples=257076 | — |
| anthropic → openai | cpu_us_per_request | 198.44375892076062 | valid | — |
| anthropic → openai | rps_per_cpu_second | 5039.211136890952 | valid | — |
| anthropic → openai | cost_window_conc | 8 | valid | — |
| anthropic → openai | cost_core_utilisation | 0.7160658692266068 | valid | — |
| anthropic → openai | cost_window_ok | 86876 | valid | — |
| anthropic → openai | cost_window_rps | 14477 | valid | — |
| anthropic → openai | cost_threads | 12 | valid | — |
| anthropic → openai | cost_nonvol_ctxt_per_request | 0.0005755329435056862 | valid | — |
| anthropic → openai | cost_majflt | 0 | valid | — |
| anthropic → openai | added_ttft_p50_us | 10646 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → openai | added_ttft_p99_us | 11159 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → openai | added_gap_p50_us | below_resolution | below_resolution; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → openai | added_gap_p99_us | 9055 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → openai | streams_sustained_fps | 77198.52773250597 | valid | — |
| anthropic → openai | streams_sustained | 1919 | valid | — |
| anthropic → openai | idle_rss_mib | 190.64453125 | valid | — |
| anthropic → openai | steady_state_rss_mib | 224.5859375 | valid | — |
| anthropic → openai | recovered_rss_mib | 224.65625 | valid | — |
| anthropic → openai | peak_rss_mib | 224.83203125 | valid | — |
| anthropic → openai | peak_rss_hwm_mib | 229.5546875 | valid | — |
| anthropic → openai | time_to_plateau_s | 66 | valid | — |
| anthropic → openai | growth_rate_mib_per_min | 0.025964471941633756 | valid | — |
| anthropic → openai | load_s | 300 | valid | — |
| anthropic → openai | shape | not_measured | not_measured | — |
| anthropic → openai | idle_shape | 0 | valid | — |
| anthropic → openai | idle_static | 0 | valid | — |
| anthropic → openai | idle_growth_rate_mib_per_min | -0.45247832091501883 | valid | — |
| anthropic → openai | throughput_rps | 11298 req/s | concurrency 4; samples 67801 | 1 |
| anthropic → openai | throughput_rps | 21698 req/s | concurrency 32; samples 130205 | 5 |
| anthropic → openai | throughput_rps | 24156 req/s | concurrency 64; samples 144943 | 10 |
| anthropic → openai | throughput_rps | 25454 req/s | concurrency 256; samples 152737 | 50 |
| anthropic → openai | throughput_rps | 25454 req/s | concurrency 256; samples 152737 | 100 |
| anthropic → openai | throughput_rps | 25454 req/s | concurrency 256; samples 152737 | unbounded |
| gemini → openai | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai-responses → openai | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| cohere → openai | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| gemini → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai-responses → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| bedrock → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| cohere → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| anthropic → gemini | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| anthropic → bedrock | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → bedrock | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| gemini → bedrock | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| bedrock → bedrock | added_latency_p50_us | 234 | valid; gateway_c1_samples=21813; direct_c1_samples=263299 | — |
| bedrock → bedrock | added_latency_p99_us | 463 | valid; gateway_c1_samples=21813; direct_c1_samples=263299 | — |
| bedrock → bedrock | gateway_c1_p99_us | 491 | valid; gateway_c1_samples=21813 | — |
| bedrock → bedrock | direct_c1_p99_us | 28 | valid; direct_c1_samples=263299 | — |
| bedrock → bedrock | cpu_us_per_request | 244.17659804983748 | valid | — |
| bedrock → bedrock | rps_per_cpu_second | 4095.396561286744 | valid | — |
| bedrock → bedrock | cost_window_conc | 8 | valid | — |
| bedrock → bedrock | cost_core_utilisation | 0.748941474283438 | valid | — |
| bedrock → bedrock | cost_window_ok | 73840 | valid | — |
| bedrock → bedrock | cost_window_rps | 12305 | valid | — |
| bedrock → bedrock | cost_threads | 12 | valid | — |
| bedrock → bedrock | cost_nonvol_ctxt_per_request | 0.0006365113759479957 | valid | — |
| bedrock → bedrock | cost_majflt | 0 | valid | — |
| bedrock → bedrock | added_ttft_p50_us | untestable | untestable | — |
| bedrock → bedrock | added_ttft_p99_us | untestable | untestable | — |
| bedrock → bedrock | added_gap_p50_us | untestable | untestable | — |
| bedrock → bedrock | added_gap_p99_us | untestable | untestable | — |
| bedrock → bedrock | streams_sustained_fps | untestable | untestable | — |
| bedrock → bedrock | streams_sustained | untestable | untestable | — |
| bedrock → bedrock | idle_rss_mib | 190.5390625 | valid | — |
| bedrock → bedrock | steady_state_rss_mib | 238.904296875 | valid | — |
| bedrock → bedrock | recovered_rss_mib | 239.44140625 | valid | — |
| bedrock → bedrock | peak_rss_mib | 240.44140625 | valid | — |
| bedrock → bedrock | peak_rss_hwm_mib | 244.83984375 | valid | — |
| bedrock → bedrock | time_to_plateau_s | not_measured | not_measured | — |
| bedrock → bedrock | growth_rate_mib_per_min | 0.37490060029269184 | valid | — |
| bedrock → bedrock | load_s | 300 | valid | — |
| bedrock → bedrock | shape | 0 | valid | — |
| bedrock → bedrock | idle_shape | 0 | valid | — |
| bedrock → bedrock | idle_static | 0 | valid | — |
| bedrock → bedrock | idle_growth_rate_mib_per_min | -0.4582365113281113 | valid | — |
| bedrock → bedrock | throughput_rps | 6434 req/s | concurrency 2; samples 38611 | 1 |
| bedrock → bedrock | throughput_rps | 18244 req/s | concurrency 32; samples 109475 | 5 |
| bedrock → bedrock | throughput_rps | 19877 req/s | concurrency 64; samples 119271 | 10 |
| bedrock → bedrock | throughput_rps | 20783 req/s | concurrency 256; samples 124713 | 50 |
| bedrock → bedrock | throughput_rps | 20783 req/s | concurrency 256; samples 124713 | 100 |
| bedrock → bedrock | throughput_rps | 20783 req/s | concurrency 256; samples 124713 | unbounded |
| cohere → bedrock | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai-responses → bedrock | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai-responses → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| bedrock → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| gemini → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| cohere → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| anthropic → cohere | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| cohere → anthropic | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| anthropic → anthropic | added_latency_p50_us | 181 | valid; gateway_c1_samples=28060; direct_c1_samples=258354 | — |
| anthropic → anthropic | added_latency_p99_us | 340 | valid; gateway_c1_samples=28060; direct_c1_samples=258354 | — |
| anthropic → anthropic | gateway_c1_p99_us | 368 | valid; gateway_c1_samples=28060 | — |
| anthropic → anthropic | direct_c1_p99_us | 28 | valid; direct_c1_samples=258354 | — |
| anthropic → anthropic | cpu_us_per_request | 184.98662543312153 | valid | — |
| anthropic → anthropic | rps_per_cpu_second | 5405.796217205613 | valid | — |
| anthropic → anthropic | cost_window_conc | 8 | valid | — |
| anthropic → anthropic | cost_core_utilisation | 0.6807669634417425 | valid | — |
| anthropic → anthropic | cost_window_ok | 88601 | valid | — |
| anthropic → anthropic | cost_window_rps | 14764 | valid | — |
| anthropic → anthropic | cost_threads | 12 | valid | — |
| anthropic → anthropic | cost_nonvol_ctxt_per_request | 0.0003724562928183655 | valid | — |
| anthropic → anthropic | cost_majflt | 0 | valid | — |
| anthropic → anthropic | added_ttft_p50_us | 10516 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → anthropic | added_ttft_p99_us | 11171 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → anthropic | added_gap_p50_us | below_resolution | below_resolution; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → anthropic | added_gap_p99_us | 8998 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → anthropic | streams_sustained_fps | 80658.75131817744 | valid | — |
| anthropic → anthropic | streams_sustained | 1844 | valid | — |
| anthropic → anthropic | idle_rss_mib | 193.1328125 | valid | — |
| anthropic → anthropic | steady_state_rss_mib | 230.0234375 | valid | — |
| anthropic → anthropic | recovered_rss_mib | 230.1953125 | valid | — |
| anthropic → anthropic | peak_rss_mib | 230.3984375 | valid | — |
| anthropic → anthropic | peak_rss_hwm_mib | 235.19140625 | valid | — |
| anthropic → anthropic | time_to_plateau_s | 66 | valid | — |
| anthropic → anthropic | growth_rate_mib_per_min | 0.1636566869781283 | valid | — |
| anthropic → anthropic | load_s | 300 | valid | — |
| anthropic → anthropic | shape | not_measured | not_measured | — |
| anthropic → anthropic | idle_shape | 0 | valid | — |
| anthropic → anthropic | idle_static | 0 | valid | — |
| anthropic → anthropic | idle_growth_rate_mib_per_min | -0.45541914018046065 | valid | — |
| anthropic → anthropic | throughput_rps | 12499 req/s | concurrency 4; samples 75005 | 1 |
| anthropic → anthropic | throughput_rps | 23800 req/s | concurrency 32; samples 142817 | 5 |
| anthropic → anthropic | throughput_rps | 26599 req/s | concurrency 64; samples 159609 | 10 |
| anthropic → anthropic | throughput_rps | 27850 req/s | concurrency 256; samples 167116 | 50 |
| anthropic → anthropic | throughput_rps | 27850 req/s | concurrency 256; samples 167116 | 100 |
| anthropic → anthropic | throughput_rps | 27850 req/s | concurrency 256; samples 167116 | unbounded |
| openai-responses → anthropic | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → anthropic | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| bedrock → anthropic | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| gemini → anthropic | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| gemini → openai-responses | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| anthropic → openai-responses | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai → openai-responses | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| cohere → openai-responses | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| bedrock → openai-responses | — | — | not_declared: Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json | — |
| openai-responses → openai-responses | added_latency_p50_us | 189 | valid; gateway_c1_samples=26997; direct_c1_samples=261509 | — |
| openai-responses → openai-responses | added_latency_p99_us | 358 | valid; gateway_c1_samples=26997; direct_c1_samples=261509 | — |
| openai-responses → openai-responses | gateway_c1_p99_us | 386 | valid; gateway_c1_samples=26997 | — |
| openai-responses → openai-responses | direct_c1_p99_us | 28 | valid; direct_c1_samples=261509 | — |
| openai-responses → openai-responses | cpu_us_per_request | 189.4636194514597 | valid | — |
| openai-responses → openai-responses | rps_per_cpu_second | 5278.058145913329 | valid | — |
| openai-responses → openai-responses | cost_window_conc | 8 | valid | — |
| openai-responses → openai-responses | cost_core_utilisation | 0.7572038580695427 | valid | — |
| openai-responses → openai-responses | cost_window_ok | 96219 | valid | — |
| openai-responses → openai-responses | cost_window_rps | 16034 | valid | — |
| openai-responses → openai-responses | cost_threads | 12 | valid | — |
| openai-responses → openai-responses | cost_nonvol_ctxt_per_request | 0.000436504224737318 | valid | — |
| openai-responses → openai-responses | cost_majflt | 0 | valid | — |
| openai-responses → openai-responses | added_ttft_p50_us | untestable | untestable | — |
| openai-responses → openai-responses | added_ttft_p99_us | untestable | untestable | — |
| openai-responses → openai-responses | added_gap_p50_us | untestable | untestable | — |
| openai-responses → openai-responses | added_gap_p99_us | untestable | untestable | — |
| openai-responses → openai-responses | streams_sustained_fps | untestable | untestable | — |
| openai-responses → openai-responses | streams_sustained | untestable | untestable | — |
| openai-responses → openai-responses | idle_rss_mib | 192.4375 | valid | — |
| openai-responses → openai-responses | steady_state_rss_mib | 228.27734375 | valid | — |
| openai-responses → openai-responses | recovered_rss_mib | 228.3125 | valid | — |
| openai-responses → openai-responses | peak_rss_mib | 228.4765625 | valid | — |
| openai-responses → openai-responses | peak_rss_hwm_mib | 233.046875 | valid | — |
| openai-responses → openai-responses | time_to_plateau_s | 66 | valid | — |
| openai-responses → openai-responses | growth_rate_mib_per_min | 0.01780973526014571 | valid | — |
| openai-responses → openai-responses | load_s | 300 | valid | — |
| openai-responses → openai-responses | shape | not_measured | not_measured | — |
| openai-responses → openai-responses | idle_shape | 0 | valid | — |
| openai-responses → openai-responses | idle_static | 0 | valid | — |
| openai-responses → openai-responses | idle_growth_rate_mib_per_min | -0.46380000104901226 | valid | — |
| openai-responses → openai-responses | throughput_rps | 11636 req/s | concurrency 4; samples 69830 | 1 |
| openai-responses → openai-responses | throughput_rps | 24247 req/s | concurrency 32; samples 145490 | 5 |
| openai-responses → openai-responses | throughput_rps | 25825 req/s | concurrency 64; samples 154951 | 10 |
| openai-responses → openai-responses | throughput_rps | 28066 req/s | concurrency 128; samples 168422 | 50 |
| openai-responses → openai-responses | throughput_rps | 28066 req/s | concurrency 128; samples 168422 | 100 |
| openai-responses → openai-responses | throughput_rps | 28066 req/s | concurrency 128; samples 168422 | 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 |
| --- | --- | --- | --- | --- | --- | --- | --- |
| bedrock → openai | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → openai | served | — | — | — | Recorded run evidence | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → openai | served | — | — | — | Recorded run evidence | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → openai | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → openai | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → openai | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| bedrock → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → gemini | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → bedrock | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → bedrock | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → bedrock | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| bedrock → bedrock | served | — | — | — | Recorded run evidence | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → bedrock | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → bedrock | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| bedrock → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → cohere | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → anthropic | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → anthropic | served | — | — | — | Recorded run evidence | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → anthropic | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → anthropic | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| bedrock → anthropic | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → anthropic | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| gemini → openai-responses | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| anthropic → openai-responses | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai → openai-responses | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| cohere → openai-responses | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| bedrock → openai-responses | not_declared | — | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 1](https://onthebench.ai/#note-1) | [recorded note 2](https://onthebench.ai/#note-2) | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |
| openai-responses → openai-responses | served | — | — | — | Recorded run evidence | — | apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38 |

**Harness commit**: `d240badd966933fa19eec0606b2d1f3945f18ad2`

**Mock SHA-256**: `99d01f709d8d9c1f49055aa265d5eb7475264d6b513911fe317aee74a71190f8`

**Load generator SHA-256**: `not recorded`

**somaxconn**: 65535

**tcp_max_syn_backlog**: 65535

## 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 → openai | 1 | {"200":1} | 0 |
| openai → openai | 2 | {"200":2} | 0 |
| openai → openai | 4 | {"200":4} | 0 |
| openai → openai | 8 | {"200":8} | 0 |
| openai → openai | 16 | {"200":16} | 0 |
| openai → openai | 32 | {"200":32} | 0 |
| openai → openai | 64 | {"200":64} | 0 |
| openai → openai | 128 | {"200":128} | 0 |
| openai → openai | 256 | {"200":256} | 0 |
| openai → openai | 512 | {"200":512} | 0 |
| openai → openai | 1024 | {"200":1024} | 0 |
| openai → openai | 2048 | {"200":2048} | 0 |
| openai → openai | 1536 | {"200":1536} | 0 |
| openai → openai | 1280 | {"200":1280} | 0 |
| openai → openai | 1408 | {"200":1408} | 0 |
| openai → openai | 1472 | {"200":1472} | 0 |
| openai → openai | 1440 | {"200":1440} | 0 |
| openai → openai | 1456 | {"200":1456} | 0 |
| openai → openai | 1464 | {"200":1464} | 0 |
| openai → openai | 1468 | {"200":1468} | 0 |
| openai → openai | 1470 | {"200":1470} | 0 |
| openai → openai | 1471 | {"200":1471} | 0 |
| openai → openai | 1470 | {"200":1470} | 0 |
| openai → openai | 1470 | {"200":1470} | 0 |
| anthropic → openai | 1 | {"200":1} | 0 |
| anthropic → openai | 2 | {"200":2} | 0 |
| anthropic → openai | 4 | {"200":4} | 0 |
| anthropic → openai | 8 | {"200":8} | 0 |
| anthropic → openai | 16 | {"200":16} | 0 |
| anthropic → openai | 32 | {"200":32} | 0 |
| anthropic → openai | 64 | {"200":64} | 0 |
| anthropic → openai | 128 | {"200":128} | 0 |
| anthropic → openai | 256 | {"200":256} | 0 |
| anthropic → openai | 512 | {"200":512} | 0 |
| anthropic → openai | 1024 | {"200":1024} | 0 |
| anthropic → openai | 2048 | {"200":2048} | 0 |
| anthropic → openai | 1536 | {"200":1536} | 0 |
| anthropic → openai | 1792 | {"200":1792} | 0 |
| anthropic → openai | 1920 | {"200":1920} | 0 |
| anthropic → openai | 1856 | {"200":1856} | 0 |
| anthropic → openai | 1888 | {"200":1888} | 0 |
| anthropic → openai | 1904 | {"200":1904} | 0 |
| anthropic → openai | 1912 | {"200":1912} | 0 |
| anthropic → openai | 1916 | {"200":1916} | 0 |
| anthropic → openai | 1918 | {"200":1918} | 0 |
| anthropic → openai | 1919 | {"200":1919} | 0 |
| anthropic → openai | 1919 | {"200":1919} | 0 |
| anthropic → openai | 1919 | {"200":1919} | 0 |
| anthropic → anthropic | 1 | {"200":1} | 0 |
| anthropic → anthropic | 2 | {"200":2} | 0 |
| anthropic → anthropic | 4 | {"200":4} | 0 |
| anthropic → anthropic | 8 | {"200":8} | 0 |
| anthropic → anthropic | 16 | {"200":16} | 0 |
| anthropic → anthropic | 32 | {"200":32} | 0 |
| anthropic → anthropic | 64 | {"200":64} | 0 |
| anthropic → anthropic | 128 | {"200":128} | 0 |
| anthropic → anthropic | 256 | {"200":256} | 0 |
| anthropic → anthropic | 512 | {"200":512} | 0 |
| anthropic → anthropic | 1024 | {"200":1024} | 0 |
| anthropic → anthropic | 2048 | {"200":2048} | 0 |
| anthropic → anthropic | 1536 | {"200":1536} | 0 |
| anthropic → anthropic | 1792 | {"200":1792} | 0 |
| anthropic → anthropic | 1920 | {"200":1920} | 0 |
| anthropic → anthropic | 1856 | {"200":1856} | 0 |
| anthropic → anthropic | 1824 | {"200":1824} | 0 |
| anthropic → anthropic | 1840 | {"200":1840} | 0 |
| anthropic → anthropic | 1848 | {"200":1848} | 0 |
| anthropic → anthropic | 1844 | {"200":1844} | 0 |
| anthropic → anthropic | 1846 | {"200":1846} | 0 |
| anthropic → anthropic | 1845 | {"200":1845} | 0 |
| anthropic → anthropic | 1844 | {"200":1844} | 0 |
| anthropic → anthropic | 1844 | {"200":1844} | 0 |

## Recorded notes

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

- Pinned APISIX 3.19.0 source 473043623c161966f3f6ef4e8f12d2d8667a12d2: ai-providers capability tables and ai-protocols/converters registry support four native chat protocol paths plus Anthropic Messages to OpenAI Chat Completions conversion. The Gemini provider uses an OpenAI-compatible endpoint, not native Gemini generateContent; provider brands are not protocol support. Five exact pairs passed local response and recorder checks. Evidence: results/diagnostics/refresh-configuration/apisix-protocol-preflight.json
- 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

- [run-apisix-cc145fc7e19bd52b5c47dc4c3db32fb0511f656418b54e198d2b3cf922523d3f](https://onthebench.ai/gateways/runs/run-apisix-cc145fc7e19bd52b5c47dc4c3db32fb0511f656418b54e198d2b3cf922523d3f)  ·  [complete raw record including RSS series](https://onthebench.ai/api/v1/runs/run-apisix-cc145fc7e19bd52b5c47dc4c3db32fb0511f656418b54e198d2b3cf922523d3f/raw.json)

## Configuration recorded by the run

```
# apisix.gen.yaml
routes:
  - id: ai-proxy-chat
    uri: /v1/chat/completions
    methods: [POST]
    plugins:
      ai-proxy:
        provider: openai-compatible
        auth: { header: { Authorization: "Bearer sk-fake-benchmark-key" } }
        options: { model: gpt-4o-mini }
        override: { endpoint: "http://127.0.0.1:8000" }
  - id: ai-proxy-responses
    uri: /v1/responses
    methods: [POST]
    plugins:
      ai-proxy:
        provider: openai-compatible
        auth: { header: { Authorization: "Bearer sk-fake-benchmark-key" } }
        options: { model: gpt-4o-mini }
        override: { endpoint: "http://127.0.0.1:8000" }
  - id: ai-proxy-messages
    uri: /v1/messages
    methods: [POST]
    plugins:
      ai-proxy:
        provider: anthropic
        auth: { header: { Authorization: "Bearer sk-fake-benchmark-key" } }
        options: { model: claude-3-5-sonnet-20241022 }
        override: { endpoint: "http://127.0.0.1:8000" }
  - id: ai-proxy-converse
    uri: /model/anthropic.claude-3-sonnet-20240229-v1:0/converse
    methods: [POST]
    plugins:
      ai-proxy:
        provider: bedrock
        provider_conf: { region: "us-east-1" }
        auth:
          aws:
            access_key_id: "AKIAMOCKACCESSKEY"
            secret_access_key: "mock-secret-access-key"
        options: { model: anthropic.claude-3-sonnet-20240229-v1:0 }
        override: { endpoint: "http://127.0.0.1:8000" }
  - id: anthropic-to-openai
    uri: /v1/messages
    priority: 10
    vars: [["http_x_bench_egress", "==", "openai"]]
    methods: [POST]
    plugins:
      ai-proxy:
        provider: openai-compatible
        auth: { header: { Authorization: "Bearer sk-fake-benchmark-key" } }
        options: { model: gpt-4o-mini }
        override: { endpoint: "http://127.0.0.1:8000" }
#END

# config.gen.yaml
deployment:
  role: data_plane
  role_data_plane:
    config_provider: yaml
apisix:
  node_listen:
    - 9080
nginx_config:
  worker_processes: 4

```

## Compare a shared protocol cell

- [APISIX vs agentgateway](https://onthebench.ai/gateways/compare/apisix-vs-agentgateway)
- [APISIX vs AISIX (api7)](https://onthebench.ai/gateways/compare/apisix-vs-aisix)
- [APISIX vs Bifrost](https://onthebench.ai/gateways/compare/apisix-vs-bifrost)
- [APISIX vs Busbar](https://onthebench.ai/gateways/compare/apisix-vs-busbar)
- [APISIX vs GoModel](https://onthebench.ai/gateways/compare/apisix-vs-gomodel)
- [APISIX vs Helicone](https://onthebench.ai/gateways/compare/apisix-vs-helicone)
- [APISIX vs Higress](https://onthebench.ai/gateways/compare/apisix-vs-higress)
- [APISIX vs Kong](https://onthebench.ai/gateways/compare/apisix-vs-kong)
- [APISIX vs LiteLLM · Python](https://onthebench.ai/gateways/compare/apisix-vs-litellm-python)
- [APISIX vs LiteLLM · Rust](https://onthebench.ai/gateways/compare/apisix-vs-litellm-rust)
- [APISIX vs One-API](https://onthebench.ai/gateways/compare/apisix-vs-one-api)
- [APISIX vs Plano](https://onthebench.ai/gateways/compare/apisix-vs-plano)
- [APISIX vs Portkey](https://onthebench.ai/gateways/compare/apisix-vs-portkey)
- [APISIX vs TensorZero](https://onthebench.ai/gateways/compare/apisix-vs-tensorzero)

## Historical measurements

- [2026-11-10T17:39:33Z — apache/apisix:3.19.0-debian@sha256:7401ce7cfd56de1b9dcecfe7d7c7e5a681bc29222767fd28391995a0936eaa38](https://onthebench.ai/gateways/runs/run-apisix-cc145fc7e19bd52b5c47dc4c3db32fb0511f656418b54e198d2b3cf922523d3f)
- [2026-08-26T19:08:41Z — apache/apisix:3.18.0-debian](https://onthebench.ai/gateways/runs/run-apisix-d68eb62b25a8380960007441ef46ed419673d0e6192cf449af6b43fb70a6e73c)
