# GoModel 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) .

Go · recorded build: enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05. Measured: 2026-11-10T17:39:52Z.

[JSON](https://onthebench.ai/api/v1/gateways/gomodel)  ·  [Markdown](https://onthebench.ai/gateways/gomodel.md)  ·  [CSV](https://onthebench.ai/gateways/data.csv?gateway=gomodel&compact=1)  ·  [Project source](https://github.com/ENTERPILOT/GOModel)  ·  [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) |
| --- | --- | --- | --- | --- |
| cohere → anthropic | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai → anthropic | added_latency_p50_us | 332 | valid; gateway_c1_samples=16334; direct_c1_samples=245181 | — |
| openai → anthropic | added_latency_p99_us | 508 | valid; gateway_c1_samples=16334; direct_c1_samples=245181 | — |
| openai → anthropic | gateway_c1_p99_us | 538 | valid; gateway_c1_samples=16334 | — |
| openai → anthropic | direct_c1_p99_us | 30 | valid; direct_c1_samples=245181 | — |
| openai → anthropic | cpu_us_per_request | 424.9108011676938 | valid | — |
| openai → anthropic | rps_per_cpu_second | 2353.4351145038167 | valid | — |
| openai → anthropic | cost_window_conc | 8 | valid | — |
| openai → anthropic | cost_core_utilisation | 0.7074297904213582 | valid | — |
| openai → anthropic | cost_window_ok | 40079 | valid | — |
| openai → anthropic | cost_window_rps | 6679 | valid | — |
| openai → anthropic | cost_threads | 13 | valid | — |
| openai → anthropic | cost_nonvol_ctxt_per_request | 0 | valid | — |
| openai → anthropic | cost_majflt | 0 | valid | — |
| openai → anthropic | added_ttft_p50_us | 482 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| openai → anthropic | added_ttft_p99_us | 635 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| openai → anthropic | added_gap_p50_us | 5 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| openai → anthropic | added_gap_p99_us | 21 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| openai → anthropic | streams_sustained_fps | 47311.46651648458 | valid | — |
| openai → anthropic | streams_sustained | 1140 | valid | — |
| openai → anthropic | idle_rss_mib | 52.9140625 | valid | — |
| openai → anthropic | steady_state_rss_mib | 87.58203125 | valid | — |
| openai → anthropic | recovered_rss_mib | 87.0859375 | valid | — |
| openai → anthropic | peak_rss_mib | 89.44140625 | valid | — |
| openai → anthropic | peak_rss_hwm_mib | 89.0703125 | valid | — |
| openai → anthropic | time_to_plateau_s | not_measured | not_measured | — |
| openai → anthropic | growth_rate_mib_per_min | -0.11046857868890064 | valid | — |
| openai → anthropic | load_s | 300 | valid | — |
| openai → anthropic | shape | 0 | valid | — |
| openai → anthropic | idle_shape | not_measured | not_measured | — |
| openai → anthropic | idle_static | 1 | valid | — |
| openai → anthropic | idle_growth_rate_mib_per_min | 0.3226255858418706 | valid | — |
| openai → anthropic | throughput_rps | 4631 req/s | concurrency 2; samples 27792 | 1 |
| openai → anthropic | throughput_rps | 6656 req/s | concurrency 8; samples 39947 | 5 |
| openai → anthropic | throughput_rps | 6820 req/s | concurrency 16; samples 40929 | 10 |
| openai → anthropic | throughput_rps | 7453 req/s | concurrency 64; samples 44724 | 50 |
| openai → anthropic | throughput_rps | 7474 req/s | concurrency 128; samples 44854 | 100 |
| openai → anthropic | throughput_rps | 7474 req/s | concurrency 128; samples 44854 | unbounded |
| openai-responses → anthropic | added_latency_p50_us | 363 | valid; gateway_c1_samples=14995; direct_c1_samples=248903 | — |
| openai-responses → anthropic | added_latency_p99_us | 639 | valid; gateway_c1_samples=14995; direct_c1_samples=248903 | — |
| openai-responses → anthropic | gateway_c1_p99_us | 669 | valid; gateway_c1_samples=14995 | — |
| openai-responses → anthropic | direct_c1_p99_us | 30 | valid; direct_c1_samples=248903 | — |
| openai-responses → anthropic | cpu_us_per_request | 571.993208450469 | valid | — |
| openai-responses → anthropic | rps_per_cpu_second | 1748.272506082725 | valid | — |
| openai-responses → anthropic | cost_window_conc | 8 | valid | — |
| openai-responses → anthropic | cost_core_utilisation | 0.8537078031502604 | valid | — |
| openai-responses → anthropic | cost_window_ok | 35927 | valid | — |
| openai-responses → anthropic | cost_window_rps | 5987 | valid | — |
| openai-responses → anthropic | cost_threads | 12 | valid | — |
| openai-responses → anthropic | cost_nonvol_ctxt_per_request | 0.009352297714810588 | valid | — |
| openai-responses → anthropic | cost_majflt | 0 | valid | — |
| openai-responses → anthropic | added_ttft_p50_us | untestable | untestable | — |
| openai-responses → anthropic | added_ttft_p99_us | untestable | untestable | — |
| openai-responses → anthropic | added_gap_p50_us | untestable | untestable | — |
| openai-responses → anthropic | added_gap_p99_us | untestable | untestable | — |
| openai-responses → anthropic | streams_sustained_fps | untestable | untestable | — |
| openai-responses → anthropic | streams_sustained | untestable | untestable | — |
| openai-responses → anthropic | idle_rss_mib | 53.15625 | valid | — |
| openai-responses → anthropic | steady_state_rss_mib | 240.859375 | valid | — |
| openai-responses → anthropic | recovered_rss_mib | 157.453125 | valid | — |
| openai-responses → anthropic | peak_rss_mib | 324.51953125 | valid | — |
| openai-responses → anthropic | peak_rss_hwm_mib | 324.09375 | valid | — |
| openai-responses → anthropic | time_to_plateau_s | not_measured | not_measured | — |
| openai-responses → anthropic | growth_rate_mib_per_min | 10.70424987337067 | valid | — |
| openai-responses → anthropic | load_s | 300 | valid | — |
| openai-responses → anthropic | shape | 0 | valid | — |
| openai-responses → anthropic | idle_shape | not_measured | not_measured | — |
| openai-responses → anthropic | idle_static | 1 | valid | — |
| openai-responses → anthropic | idle_growth_rate_mib_per_min | 0.35096701211601694 | valid | — |
| openai-responses → anthropic | throughput_rps | 4280 req/s | concurrency 2; samples 25689 | 1 |
| openai-responses → anthropic | throughput_rps | 5982 req/s | concurrency 8; samples 35897 | 5 |
| openai-responses → anthropic | throughput_rps | 5982 req/s | concurrency 8; samples 35897 | 10 |
| openai-responses → anthropic | throughput_rps | 6343 req/s | concurrency 64; samples 38060 | 50 |
| openai-responses → anthropic | throughput_rps | 6343 req/s | concurrency 64; samples 38060 | 100 |
| openai-responses → anthropic | throughput_rps | 6343 req/s | concurrency 64; samples 38060 | unbounded |
| bedrock → anthropic | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| gemini → anthropic | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| anthropic → anthropic | added_latency_p50_us | 386 | valid; gateway_c1_samples=14086; direct_c1_samples=240685 | — |
| anthropic → anthropic | added_latency_p99_us | 722 | valid; gateway_c1_samples=14086; direct_c1_samples=240685 | — |
| anthropic → anthropic | gateway_c1_p99_us | 753 | valid; gateway_c1_samples=14086 | — |
| anthropic → anthropic | direct_c1_p99_us | 31 | valid; direct_c1_samples=240685 | — |
| anthropic → anthropic | cpu_us_per_request | 498.8399071925754 | valid | — |
| anthropic → anthropic | rps_per_cpu_second | 2004.6511627906975 | valid | — |
| anthropic → anthropic | cost_window_conc | 8 | valid | — |
| anthropic → anthropic | cost_core_utilisation | 0.785925521214345 | valid | — |
| anthropic → anthropic | cost_window_ok | 37928 | valid | — |
| anthropic → anthropic | cost_window_rps | 6320 | valid | — |
| anthropic → anthropic | cost_threads | 13 | valid | — |
| anthropic → anthropic | cost_nonvol_ctxt_per_request | 0 | valid | — |
| anthropic → anthropic | cost_majflt | 0 | valid | — |
| anthropic → anthropic | added_ttft_p50_us | 485 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → anthropic | added_ttft_p99_us | 1078 | 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 | below_resolution | below_resolution; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → anthropic | streams_sustained_fps | 64695.44352492808 | valid | — |
| anthropic → anthropic | streams_sustained | 1559 | valid | — |
| anthropic → anthropic | idle_rss_mib | 52.9609375 | valid | — |
| anthropic → anthropic | steady_state_rss_mib | 86.8515625 | valid | — |
| anthropic → anthropic | recovered_rss_mib | 87.08203125 | valid | — |
| anthropic → anthropic | peak_rss_mib | 88.515625 | valid | — |
| anthropic → anthropic | peak_rss_hwm_mib | 88.10546875 | valid | — |
| anthropic → anthropic | time_to_plateau_s | not_measured | not_measured | — |
| anthropic → anthropic | growth_rate_mib_per_min | -0.2543254131118377 | valid | — |
| anthropic → anthropic | load_s | 300 | valid | — |
| anthropic → anthropic | shape | 0 | valid | — |
| anthropic → anthropic | idle_shape | not_measured | not_measured | — |
| anthropic → anthropic | idle_static | 1 | valid | — |
| anthropic → anthropic | idle_growth_rate_mib_per_min | 0.3148156669010287 | valid | — |
| anthropic → anthropic | throughput_rps | 2351 req/s | concurrency 1; samples 14110 | 1 |
| anthropic → anthropic | throughput_rps | 6462 req/s | concurrency 8; samples 38778 | 5 |
| anthropic → anthropic | throughput_rps | 6654 req/s | concurrency 16; samples 39931 | 10 |
| anthropic → anthropic | throughput_rps | 7222 req/s | concurrency 64; samples 43337 | 50 |
| anthropic → anthropic | throughput_rps | 7237 req/s | concurrency 128; samples 43430 | 100 |
| anthropic → anthropic | throughput_rps | 7237 req/s | concurrency 128; samples 43430 | unbounded |
| bedrock → openai | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| cohere → openai | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai-responses → openai | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| gemini → openai | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| anthropic → openai | added_latency_p50_us | 348 | valid; gateway_c1_samples=15595; direct_c1_samples=242839 | — |
| anthropic → openai | added_latency_p99_us | 531 | valid; gateway_c1_samples=15595; direct_c1_samples=242839 | — |
| anthropic → openai | gateway_c1_p99_us | 561 | valid; gateway_c1_samples=15595 | — |
| anthropic → openai | direct_c1_p99_us | 30 | valid; direct_c1_samples=242839 | — |
| anthropic → openai | cpu_us_per_request | 449.57994470837696 | valid | — |
| anthropic → openai | rps_per_cpu_second | 2224.2985074626868 | valid | — |
| anthropic → openai | cost_window_conc | 8 | valid | — |
| anthropic → openai | cost_core_utilisation | 0.695593380601896 | valid | — |
| anthropic → openai | cost_window_ok | 37257 | valid | — |
| anthropic → openai | cost_window_rps | 6208 | valid | — |
| anthropic → openai | cost_threads | 13 | valid | — |
| anthropic → openai | cost_nonvol_ctxt_per_request | 0.0055291623050701885 | valid | — |
| anthropic → openai | cost_majflt | 0 | valid | — |
| anthropic → openai | added_ttft_p50_us | 482 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| anthropic → openai | added_ttft_p99_us | 681 | 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 | 14 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| anthropic → openai | streams_sustained_fps | 61069.630667374906 | valid | — |
| anthropic → openai | streams_sustained | 1536 | valid | — |
| anthropic → openai | idle_rss_mib | 53.23828125 | valid | — |
| anthropic → openai | steady_state_rss_mib | 87.93359375 | valid | — |
| anthropic → openai | recovered_rss_mib | 89.765625 | valid | — |
| anthropic → openai | peak_rss_mib | 90.3125 | valid | — |
| anthropic → openai | peak_rss_hwm_mib | 90 | valid | — |
| anthropic → openai | time_to_plateau_s | not_measured | not_measured | — |
| anthropic → openai | growth_rate_mib_per_min | 0.027332179074735487 | valid | — |
| anthropic → openai | load_s | 300 | valid | — |
| anthropic → openai | shape | 0 | valid | — |
| anthropic → openai | idle_shape | not_measured | not_measured | — |
| anthropic → openai | idle_static | 1 | valid | — |
| anthropic → openai | idle_growth_rate_mib_per_min | 0.19131230185335213 | valid | — |
| anthropic → openai | throughput_rps | 4318 req/s | concurrency 2; samples 25914 | 1 |
| anthropic → openai | throughput_rps | 6214 req/s | concurrency 8; samples 37292 | 5 |
| anthropic → openai | throughput_rps | 6384 req/s | concurrency 16; samples 38313 | 10 |
| anthropic → openai | throughput_rps | 6988 req/s | concurrency 64; samples 41937 | 50 |
| anthropic → openai | throughput_rps | 7071 req/s | concurrency 128; samples 42435 | 100 |
| anthropic → openai | throughput_rps | 7071 req/s | concurrency 128; samples 42435 | unbounded |
| openai → openai | added_latency_p50_us | 343 | valid; gateway_c1_samples=15837; direct_c1_samples=239533 | — |
| openai → openai | added_latency_p99_us | 565 | valid; gateway_c1_samples=15837; direct_c1_samples=239533 | — |
| openai → openai | gateway_c1_p99_us | 596 | valid; gateway_c1_samples=15837 | — |
| openai → openai | direct_c1_p99_us | 31 | valid; direct_c1_samples=239533 | — |
| openai → openai | cpu_us_per_request | 439.1503284680862 | valid | — |
| openai → openai | rps_per_cpu_second | 2277.1245634458674 | valid | — |
| openai → openai | cost_window_conc | 8 | valid | — |
| openai → openai | cost_core_utilisation | 0.7135981462725213 | valid | — |
| openai → openai | cost_window_ok | 39121 | valid | — |
| openai → openai | cost_window_rps | 6519 | valid | — |
| openai → openai | cost_threads | 14 | valid | — |
| openai → openai | cost_nonvol_ctxt_per_request | 0 | valid | — |
| openai → openai | cost_majflt | 0 | valid | — |
| openai → openai | added_ttft_p50_us | 479 | valid; ttft_gw_samples=100; ttft_direct_samples=100 | — |
| openai → openai | added_ttft_p99_us | 601 | 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 | 26 | valid; gap_gw_samples=63; gap_direct_samples=63 | — |
| openai → openai | streams_sustained_fps | 50390.08056872798 | valid | — |
| openai → openai | streams_sustained | 1208 | valid | — |
| openai → openai | idle_rss_mib | 53.00390625 | valid | — |
| openai → openai | steady_state_rss_mib | 87.26171875 | valid | — |
| openai → openai | recovered_rss_mib | 90.45703125 | valid | — |
| openai → openai | peak_rss_mib | 90.5078125 | valid | — |
| openai → openai | peak_rss_hwm_mib | 89.015625 | valid | — |
| openai → openai | time_to_plateau_s | not_measured | not_measured | — |
| openai → openai | growth_rate_mib_per_min | -0.33822817505778513 | valid | — |
| openai → openai | load_s | 300 | valid | — |
| openai → openai | shape | 0 | valid | — |
| openai → openai | idle_shape | not_measured | not_measured | — |
| openai → openai | idle_static | 1 | valid | — |
| openai → openai | idle_growth_rate_mib_per_min | 0.41083926712820346 | valid | — |
| openai → openai | throughput_rps | 4555 req/s | concurrency 2; samples 27335 | 1 |
| openai → openai | throughput_rps | 6685 req/s | concurrency 8; samples 40117 | 5 |
| openai → openai | throughput_rps | 6862 req/s | concurrency 16; samples 41180 | 10 |
| openai → openai | throughput_rps | 7415 req/s | concurrency 64; samples 44496 | 50 |
| openai → openai | throughput_rps | 7475 req/s | concurrency 128; samples 44859 | 100 |
| openai → openai | throughput_rps | 7475 req/s | concurrency 128; samples 44859 | unbounded |
| gemini → cohere | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai-responses → cohere | added_latency_p50_us | 353 | valid; gateway_c1_samples=15319; direct_c1_samples=247454 | — |
| openai-responses → cohere | added_latency_p99_us | 600 | valid; gateway_c1_samples=15319; direct_c1_samples=247454 | — |
| openai-responses → cohere | gateway_c1_p99_us | 630 | valid; gateway_c1_samples=15319 | — |
| openai-responses → cohere | direct_c1_p99_us | 30 | valid; direct_c1_samples=247454 | — |
| openai-responses → cohere | cpu_us_per_request | 569.9209901099508 | valid | — |
| openai-responses → cohere | rps_per_cpu_second | 1754.6291808046535 | valid | — |
| openai-responses → cohere | cost_window_conc | 8 | valid | — |
| openai-responses → cohere | cost_core_utilisation | 0.8569936858107657 | valid | — |
| openai-responses → cohere | cost_window_ok | 36198 | valid | — |
| openai-responses → cohere | cost_window_rps | 6032 | valid | — |
| openai-responses → cohere | cost_threads | 12 | valid | — |
| openai-responses → cohere | cost_nonvol_ctxt_per_request | 0.013011768606000331 | valid | — |
| openai-responses → cohere | cost_majflt | 0 | valid | — |
| openai-responses → cohere | added_ttft_p50_us | untestable | untestable | — |
| openai-responses → cohere | added_ttft_p99_us | untestable | untestable | — |
| openai-responses → cohere | added_gap_p50_us | untestable | untestable | — |
| openai-responses → cohere | added_gap_p99_us | untestable | untestable | — |
| openai-responses → cohere | streams_sustained_fps | untestable | untestable | — |
| openai-responses → cohere | streams_sustained | untestable | untestable | — |
| openai-responses → cohere | idle_rss_mib | 53.12109375 | valid | — |
| openai-responses → cohere | steady_state_rss_mib | 229.89453125 | valid | — |
| openai-responses → cohere | recovered_rss_mib | 145.29296875 | valid | — |
| openai-responses → cohere | peak_rss_mib | 316.98046875 | valid | — |
| openai-responses → cohere | peak_rss_hwm_mib | 316.4453125 | valid | — |
| openai-responses → cohere | time_to_plateau_s | not_measured | not_measured | — |
| openai-responses → cohere | growth_rate_mib_per_min | -59.167562410516275 | valid | — |
| openai-responses → cohere | load_s | 300 | valid | — |
| openai-responses → cohere | shape | -1 | valid | — |
| openai-responses → cohere | idle_shape | not_measured | not_measured | — |
| openai-responses → cohere | idle_static | 1 | valid | — |
| openai-responses → cohere | idle_growth_rate_mib_per_min | 0.3443264303539636 | valid | — |
| openai-responses → cohere | throughput_rps | 4322 req/s | concurrency 2; samples 25940 | 1 |
| openai-responses → cohere | throughput_rps | 6006 req/s | concurrency 8; samples 36046 | 5 |
| openai-responses → cohere | throughput_rps | 6006 req/s | concurrency 8; samples 36046 | 10 |
| openai-responses → cohere | throughput_rps | 6311 req/s | concurrency 64; samples 37871 | 50 |
| openai-responses → cohere | throughput_rps | 6311 req/s | concurrency 64; samples 37871 | 100 |
| openai-responses → cohere | throughput_rps | 6311 req/s | concurrency 64; samples 37871 | unbounded |
| anthropic → cohere | added_latency_p50_us | 329 | valid; gateway_c1_samples=16446; direct_c1_samples=244026 | — |
| anthropic → cohere | added_latency_p99_us | 513 | valid; gateway_c1_samples=16446; direct_c1_samples=244026 | — |
| anthropic → cohere | gateway_c1_p99_us | 543 | valid; gateway_c1_samples=16446 | — |
| anthropic → cohere | direct_c1_p99_us | 30 | valid; direct_c1_samples=244026 | — |
| anthropic → cohere | cpu_us_per_request | 426.17841838722245 | valid | — |
| anthropic → cohere | rps_per_cpu_second | 2346.435100548446 | valid | — |
| anthropic → cohere | cost_window_conc | 8 | valid | — |
| anthropic → cohere | cost_core_utilisation | 0.6817240720590323 | valid | — |
| anthropic → cohere | cost_window_ok | 38505 | valid | — |
| anthropic → cohere | cost_window_rps | 6416 | valid | — |
| anthropic → cohere | cost_threads | 12 | valid | — |
| anthropic → cohere | cost_nonvol_ctxt_per_request | 0.005609661082976237 | valid | — |
| anthropic → cohere | cost_majflt | 0 | valid | — |
| anthropic → cohere | added_ttft_p50_us | untestable | untestable | — |
| anthropic → cohere | added_ttft_p99_us | untestable | untestable | — |
| anthropic → cohere | added_gap_p50_us | untestable | untestable | — |
| anthropic → cohere | added_gap_p99_us | untestable | untestable | — |
| anthropic → cohere | streams_sustained_fps | untestable | untestable | — |
| anthropic → cohere | streams_sustained | untestable | untestable | — |
| anthropic → cohere | idle_rss_mib | 53.11328125 | valid | — |
| anthropic → cohere | steady_state_rss_mib | 87.6015625 | valid | — |
| anthropic → cohere | recovered_rss_mib | 88.75390625 | valid | — |
| anthropic → cohere | peak_rss_mib | 89.65234375 | valid | — |
| anthropic → cohere | peak_rss_hwm_mib | 89.16015625 | valid | — |
| anthropic → cohere | time_to_plateau_s | not_measured | not_measured | — |
| anthropic → cohere | growth_rate_mib_per_min | 0.11396281504661937 | valid | — |
| anthropic → cohere | load_s | 300 | valid | — |
| anthropic → cohere | shape | 0 | valid | — |
| anthropic → cohere | idle_shape | not_measured | not_measured | — |
| anthropic → cohere | idle_static | 1 | valid | — |
| anthropic → cohere | idle_growth_rate_mib_per_min | 0.2512557476501924 | valid | — |
| anthropic → cohere | throughput_rps | 4543 req/s | concurrency 2; samples 27264 | 1 |
| anthropic → cohere | throughput_rps | 6320 req/s | concurrency 8; samples 37931 | 5 |
| anthropic → cohere | throughput_rps | 6500 req/s | concurrency 16; samples 39010 | 10 |
| anthropic → cohere | throughput_rps | 7107 req/s | concurrency 64; samples 42647 | 50 |
| anthropic → cohere | throughput_rps | 7172 req/s | concurrency 128; samples 43039 | 100 |
| anthropic → cohere | throughput_rps | 7172 req/s | concurrency 128; samples 43039 | unbounded |
| cohere → cohere | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| bedrock → cohere | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai → cohere | added_latency_p50_us | 335 | valid; gateway_c1_samples=16166; direct_c1_samples=236668 | — |
| openai → cohere | added_latency_p99_us | 515 | valid; gateway_c1_samples=16166; direct_c1_samples=236668 | — |
| openai → cohere | gateway_c1_p99_us | 546 | valid; gateway_c1_samples=16166 | — |
| openai → cohere | direct_c1_p99_us | 31 | valid; direct_c1_samples=236668 | — |
| openai → cohere | cpu_us_per_request | 418.7076923076923 | valid | — |
| openai → cohere | rps_per_cpu_second | 2388.3009994121103 | valid | — |
| openai → cohere | cost_window_conc | 8 | valid | — |
| openai → cohere | cost_core_utilisation | 0.7066179481631804 | valid | — |
| openai → cohere | cost_window_ok | 40625 | valid | — |
| openai → cohere | cost_window_rps | 6769 | valid | — |
| openai → cohere | cost_threads | 12 | valid | — |
| openai → cohere | cost_nonvol_ctxt_per_request | 0.005267692307692307 | valid | — |
| openai → cohere | cost_majflt | 0 | valid | — |
| openai → cohere | added_ttft_p50_us | untestable | untestable | — |
| openai → cohere | added_ttft_p99_us | untestable | untestable | — |
| openai → cohere | added_gap_p50_us | untestable | untestable | — |
| openai → cohere | added_gap_p99_us | untestable | untestable | — |
| openai → cohere | streams_sustained_fps | untestable | untestable | — |
| openai → cohere | streams_sustained | untestable | untestable | — |
| openai → cohere | idle_rss_mib | 52.87890625 | valid | — |
| openai → cohere | steady_state_rss_mib | 87.42578125 | valid | — |
| openai → cohere | recovered_rss_mib | 90.28125 | valid | — |
| openai → cohere | peak_rss_mib | 90.28125 | valid | — |
| openai → cohere | peak_rss_hwm_mib | 88.984375 | valid | — |
| openai → cohere | time_to_plateau_s | not_measured | not_measured | — |
| openai → cohere | growth_rate_mib_per_min | 0.20381370094580964 | valid | — |
| openai → cohere | load_s | 300 | valid | — |
| openai → cohere | shape | 0 | valid | — |
| openai → cohere | idle_shape | not_measured | not_measured | — |
| openai → cohere | idle_static | 1 | valid | — |
| openai → cohere | idle_growth_rate_mib_per_min | 0.25987586900840676 | valid | — |
| openai → cohere | throughput_rps | 4690 req/s | concurrency 2; samples 28153 | 1 |
| openai → cohere | throughput_rps | 6742 req/s | concurrency 8; samples 40458 | 5 |
| openai → cohere | throughput_rps | 6886 req/s | concurrency 16; samples 41331 | 10 |
| openai → cohere | throughput_rps | 7538 req/s | concurrency 64; samples 45232 | 50 |
| openai → cohere | throughput_rps | 7549 req/s | concurrency 128; samples 45303 | 100 |
| openai → cohere | throughput_rps | 7549 req/s | concurrency 128; samples 45303 | unbounded |
| anthropic → bedrock | added_latency_p50_us | 452 | valid; gateway_c1_samples=12219; direct_c1_samples=247325 | — |
| anthropic → bedrock | added_latency_p99_us | 774 | valid; gateway_c1_samples=12219; direct_c1_samples=247325 | — |
| anthropic → bedrock | gateway_c1_p99_us | 804 | valid; gateway_c1_samples=12219 | — |
| anthropic → bedrock | direct_c1_p99_us | 30 | valid; direct_c1_samples=247325 | — |
| anthropic → bedrock | cpu_us_per_request | 553.5889579676949 | valid | — |
| anthropic → bedrock | rps_per_cpu_second | 1806.3944116066632 | valid | — |
| anthropic → bedrock | cost_window_conc | 8 | valid | — |
| anthropic → bedrock | cost_core_utilisation | 0.7731420943510743 | valid | — |
| anthropic → bedrock | cost_window_ok | 33617 | valid | — |
| anthropic → bedrock | cost_window_rps | 5602 | valid | — |
| anthropic → bedrock | cost_threads | 12 | valid | — |
| anthropic → bedrock | cost_nonvol_ctxt_per_request | 0.006990510753487818 | valid | — |
| anthropic → bedrock | cost_majflt | 0 | valid | — |
| anthropic → bedrock | added_ttft_p50_us | untestable | untestable | — |
| anthropic → bedrock | added_ttft_p99_us | untestable | untestable | — |
| anthropic → bedrock | added_gap_p50_us | untestable | untestable | — |
| anthropic → bedrock | added_gap_p99_us | untestable | untestable | — |
| anthropic → bedrock | streams_sustained_fps | untestable | untestable | — |
| anthropic → bedrock | streams_sustained | untestable | untestable | — |
| anthropic → bedrock | idle_rss_mib | 52.88671875 | valid | — |
| anthropic → bedrock | steady_state_rss_mib | 91.16796875 | valid | — |
| anthropic → bedrock | recovered_rss_mib | 91.66796875 | valid | — |
| anthropic → bedrock | peak_rss_mib | 93.25390625 | valid | — |
| anthropic → bedrock | peak_rss_hwm_mib | 92.7109375 | valid | — |
| anthropic → bedrock | time_to_plateau_s | not_measured | not_measured | — |
| anthropic → bedrock | growth_rate_mib_per_min | -0.046980188503311364 | valid | — |
| anthropic → bedrock | load_s | 300 | valid | — |
| anthropic → bedrock | shape | 0 | valid | — |
| anthropic → bedrock | idle_shape | 0 | valid | — |
| anthropic → bedrock | idle_static | 0 | valid | — |
| anthropic → bedrock | idle_growth_rate_mib_per_min | 0.38419409473514465 | valid | — |
| anthropic → bedrock | throughput_rps | 2020 req/s | concurrency 1; samples 12127 | 1 |
| anthropic → bedrock | throughput_rps | 5567 req/s | concurrency 8; samples 33409 | 5 |
| anthropic → bedrock | throughput_rps | 5567 req/s | concurrency 8; samples 33409 | 10 |
| anthropic → bedrock | throughput_rps | 6093 req/s | concurrency 64; samples 36562 | 50 |
| anthropic → bedrock | throughput_rps | 6134 req/s | concurrency 128; samples 36811 | 100 |
| anthropic → bedrock | throughput_rps | 6134 req/s | concurrency 128; samples 36811 | unbounded |
| cohere → bedrock | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| gemini → bedrock | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| bedrock → bedrock | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai-responses → bedrock | added_latency_p50_us | 471 | valid; gateway_c1_samples=11671; direct_c1_samples=247559 | — |
| openai-responses → bedrock | added_latency_p99_us | 809 | valid; gateway_c1_samples=11671; direct_c1_samples=247559 | — |
| openai-responses → bedrock | gateway_c1_p99_us | 839 | valid; gateway_c1_samples=11671 | — |
| openai-responses → bedrock | direct_c1_p99_us | 30 | valid; direct_c1_samples=247559 | — |
| openai-responses → bedrock | cpu_us_per_request | 674.6747391472369 | valid | — |
| openai-responses → bedrock | rps_per_cpu_second | 1482.1957040572793 | valid | — |
| openai-responses → bedrock | cost_window_conc | 8 | valid | — |
| openai-responses → bedrock | cost_core_utilisation | 0.870399681280061 | valid | — |
| openai-responses → bedrock | cost_window_ok | 31052 | valid | — |
| openai-responses → bedrock | cost_window_rps | 5174 | valid | — |
| openai-responses → bedrock | cost_threads | 13 | valid | — |
| openai-responses → bedrock | cost_nonvol_ctxt_per_request | 0.023605564858946285 | valid | — |
| openai-responses → bedrock | cost_majflt | 0 | valid | — |
| openai-responses → bedrock | added_ttft_p50_us | untestable | untestable | — |
| openai-responses → bedrock | added_ttft_p99_us | untestable | untestable | — |
| openai-responses → bedrock | added_gap_p50_us | untestable | untestable | — |
| openai-responses → bedrock | added_gap_p99_us | untestable | untestable | — |
| openai-responses → bedrock | streams_sustained_fps | untestable | untestable | — |
| openai-responses → bedrock | streams_sustained | untestable | untestable | — |
| openai-responses → bedrock | idle_rss_mib | 53.0703125 | valid | — |
| openai-responses → bedrock | steady_state_rss_mib | 193.3359375 | valid | — |
| openai-responses → bedrock | recovered_rss_mib | 130.703125 | valid | — |
| openai-responses → bedrock | peak_rss_mib | 276.3125 | valid | — |
| openai-responses → bedrock | peak_rss_hwm_mib | 275.9609375 | valid | — |
| openai-responses → bedrock | time_to_plateau_s | not_measured | not_measured | — |
| openai-responses → bedrock | growth_rate_mib_per_min | -105.13461029863961 | valid | — |
| openai-responses → bedrock | load_s | 300 | valid | — |
| openai-responses → bedrock | shape | -1 | valid | — |
| openai-responses → bedrock | idle_shape | not_measured | not_measured | — |
| openai-responses → bedrock | idle_static | 1 | valid | — |
| openai-responses → bedrock | idle_growth_rate_mib_per_min | 0.27690046602973994 | valid | — |
| openai-responses → bedrock | throughput_rps | 1945 req/s | concurrency 1; samples 11672 | 1 |
| openai-responses → bedrock | throughput_rps | 4809 req/s | concurrency 4; samples 28860 | 5 |
| openai-responses → bedrock | throughput_rps | 5149 req/s | concurrency 8; samples 30902 | 10 |
| openai-responses → bedrock | throughput_rps | 5409 req/s | concurrency 64; samples 32459 | 50 |
| openai-responses → bedrock | throughput_rps | 5409 req/s | concurrency 64; samples 32459 | 100 |
| openai-responses → bedrock | throughput_rps | 5409 req/s | concurrency 64; samples 32459 | unbounded |
| openai → bedrock | added_latency_p50_us | 450 | valid; gateway_c1_samples=12265; direct_c1_samples=239336 | — |
| openai → bedrock | added_latency_p99_us | 756 | valid; gateway_c1_samples=12265; direct_c1_samples=239336 | — |
| openai → bedrock | gateway_c1_p99_us | 787 | valid; gateway_c1_samples=12265 | — |
| openai → bedrock | direct_c1_p99_us | 31 | valid; direct_c1_samples=239336 | — |
| openai → bedrock | cpu_us_per_request | 545.4287739192063 | valid | — |
| openai → bedrock | rps_per_cpu_second | 1833.4199584199587 | valid | — |
| openai → bedrock | cost_window_conc | 8 | valid | — |
| openai → bedrock | cost_core_utilisation | 0.7992734367170747 | valid | — |
| openai → bedrock | cost_window_ok | 35275 | valid | — |
| openai → bedrock | cost_window_rps | 5878 | valid | — |
| openai → bedrock | cost_threads | 12 | valid | — |
| openai → bedrock | cost_nonvol_ctxt_per_request | 0 | valid | — |
| openai → bedrock | cost_majflt | 0 | valid | — |
| openai → bedrock | added_ttft_p50_us | untestable | untestable | — |
| openai → bedrock | added_ttft_p99_us | untestable | untestable | — |
| openai → bedrock | added_gap_p50_us | untestable | untestable | — |
| openai → bedrock | added_gap_p99_us | untestable | untestable | — |
| openai → bedrock | streams_sustained_fps | untestable | untestable | — |
| openai → bedrock | streams_sustained | untestable | untestable | — |
| openai → bedrock | idle_rss_mib | 52.40234375 | valid | — |
| openai → bedrock | steady_state_rss_mib | 91.28515625 | valid | — |
| openai → bedrock | recovered_rss_mib | 88.5234375 | valid | — |
| openai → bedrock | peak_rss_mib | 92.87890625 | valid | — |
| openai → bedrock | peak_rss_hwm_mib | 92.6640625 | valid | — |
| openai → bedrock | time_to_plateau_s | not_measured | not_measured | — |
| openai → bedrock | growth_rate_mib_per_min | 0.28900171285631676 | valid | — |
| openai → bedrock | load_s | 300 | valid | — |
| openai → bedrock | shape | 0 | valid | — |
| openai → bedrock | idle_shape | not_measured | not_measured | — |
| openai → bedrock | idle_static | 1 | valid | — |
| openai → bedrock | idle_growth_rate_mib_per_min | 0.29023336794601473 | valid | — |
| openai → bedrock | throughput_rps | 2045 req/s | concurrency 1; samples 12274 | 1 |
| openai → bedrock | throughput_rps | 5862 req/s | concurrency 8; samples 35177 | 5 |
| openai → bedrock | throughput_rps | 5862 req/s | concurrency 8; samples 35177 | 10 |
| openai → bedrock | throughput_rps | 6347 req/s | concurrency 64; samples 38087 | 50 |
| openai → bedrock | throughput_rps | 6361 req/s | concurrency 128; samples 38177 | 100 |
| openai → bedrock | throughput_rps | 6361 req/s | concurrency 128; samples 38177 | unbounded |
| cohere → gemini | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| bedrock → gemini | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| gemini → gemini | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| anthropic → gemini | added_latency_p50_us | 336 | valid; gateway_c1_samples=16147; direct_c1_samples=237693 | — |
| anthropic → gemini | added_latency_p99_us | 499 | valid; gateway_c1_samples=16147; direct_c1_samples=237693 | — |
| anthropic → gemini | gateway_c1_p99_us | 530 | valid; gateway_c1_samples=16147 | — |
| anthropic → gemini | direct_c1_p99_us | 31 | valid; direct_c1_samples=237693 | — |
| anthropic → gemini | cpu_us_per_request | 432.495001578449 | valid | — |
| anthropic → gemini | rps_per_cpu_second | 2312.1654501216544 | valid | — |
| anthropic → gemini | cost_window_conc | 8 | valid | — |
| anthropic → gemini | cost_core_utilisation | 0.68298447343346 | valid | — |
| anthropic → gemini | cost_window_ok | 38012 | valid | — |
| anthropic → gemini | cost_window_rps | 6334 | valid | — |
| anthropic → gemini | cost_threads | 12 | valid | — |
| anthropic → gemini | cost_nonvol_ctxt_per_request | 0 | valid | — |
| anthropic → gemini | cost_majflt | 0 | valid | — |
| anthropic → gemini | added_ttft_p50_us | untestable | untestable | — |
| anthropic → gemini | added_ttft_p99_us | untestable | untestable | — |
| anthropic → gemini | added_gap_p50_us | untestable | untestable | — |
| anthropic → gemini | added_gap_p99_us | untestable | untestable | — |
| anthropic → gemini | streams_sustained_fps | untestable | untestable | — |
| anthropic → gemini | streams_sustained | untestable | untestable | — |
| anthropic → gemini | idle_rss_mib | 53.17578125 | valid | — |
| anthropic → gemini | steady_state_rss_mib | 88.13671875 | valid | — |
| anthropic → gemini | recovered_rss_mib | 88.40625 | valid | — |
| anthropic → gemini | peak_rss_mib | 89.99609375 | valid | — |
| anthropic → gemini | peak_rss_hwm_mib | 89.93359375 | valid | — |
| anthropic → gemini | time_to_plateau_s | not_measured | not_measured | — |
| anthropic → gemini | growth_rate_mib_per_min | 0.2500767982750217 | valid | — |
| anthropic → gemini | load_s | 300 | valid | — |
| anthropic → gemini | shape | 0 | valid | — |
| anthropic → gemini | idle_shape | 0 | valid | — |
| anthropic → gemini | idle_static | 0 | valid | — |
| anthropic → gemini | idle_growth_rate_mib_per_min | 0.5579577971970239 | valid | — |
| anthropic → gemini | throughput_rps | 4493 req/s | concurrency 2; samples 26962 | 1 |
| anthropic → gemini | throughput_rps | 6289 req/s | concurrency 8; samples 37743 | 5 |
| anthropic → gemini | throughput_rps | 6450 req/s | concurrency 16; samples 38708 | 10 |
| anthropic → gemini | throughput_rps | 7092 req/s | concurrency 64; samples 42558 | 50 |
| anthropic → gemini | throughput_rps | 7137 req/s | concurrency 128; samples 42837 | 100 |
| anthropic → gemini | throughput_rps | 7137 req/s | concurrency 128; samples 42837 | unbounded |
| openai-responses → gemini | added_latency_p50_us | 353 | valid; gateway_c1_samples=15337; direct_c1_samples=250993 | — |
| openai-responses → gemini | added_latency_p99_us | 599 | valid; gateway_c1_samples=15337; direct_c1_samples=250993 | — |
| openai-responses → gemini | gateway_c1_p99_us | 629 | valid; gateway_c1_samples=15337 | — |
| openai-responses → gemini | direct_c1_p99_us | 30 | valid; direct_c1_samples=250993 | — |
| openai-responses → gemini | cpu_us_per_request | 571.7538598245029 | valid | — |
| openai-responses → gemini | rps_per_cpu_second | 1749.0043710539096 | valid | — |
| openai-responses → gemini | cost_window_conc | 8 | valid | — |
| openai-responses → gemini | cost_core_utilisation | 0.8553760435685774 | valid | — |
| openai-responses → gemini | cost_window_ok | 36012 | valid | — |
| openai-responses → gemini | cost_window_rps | 6001 | valid | — |
| openai-responses → gemini | cost_threads | 12 | valid | — |
| openai-responses → gemini | cost_nonvol_ctxt_per_request | 0.01721648339442408 | valid | — |
| openai-responses → gemini | cost_majflt | 0 | valid | — |
| openai-responses → gemini | added_ttft_p50_us | untestable | untestable | — |
| openai-responses → gemini | added_ttft_p99_us | untestable | untestable | — |
| openai-responses → gemini | added_gap_p50_us | untestable | untestable | — |
| openai-responses → gemini | added_gap_p99_us | untestable | untestable | — |
| openai-responses → gemini | streams_sustained_fps | untestable | untestable | — |
| openai-responses → gemini | streams_sustained | untestable | untestable | — |
| openai-responses → gemini | idle_rss_mib | 53.44921875 | valid | — |
| openai-responses → gemini | steady_state_rss_mib | 280.70703125 | valid | — |
| openai-responses → gemini | recovered_rss_mib | 149.56640625 | valid | — |
| openai-responses → gemini | peak_rss_mib | 371.6796875 | valid | — |
| openai-responses → gemini | peak_rss_hwm_mib | 371.296875 | valid | — |
| openai-responses → gemini | time_to_plateau_s | not_measured | not_measured | — |
| openai-responses → gemini | growth_rate_mib_per_min | 70.65268746342376 | valid | — |
| openai-responses → gemini | load_s | 300 | valid | — |
| openai-responses → gemini | shape | 1 | valid | — |
| openai-responses → gemini | idle_shape | 0 | valid | — |
| openai-responses → gemini | idle_static | 0 | valid | — |
| openai-responses → gemini | idle_growth_rate_mib_per_min | 0.6650826205323993 | valid | — |
| openai-responses → gemini | throughput_rps | 4292 req/s | concurrency 2; samples 25758 | 1 |
| openai-responses → gemini | throughput_rps | 5944 req/s | concurrency 8; samples 35670 | 5 |
| openai-responses → gemini | throughput_rps | 5944 req/s | concurrency 8; samples 35670 | 10 |
| openai-responses → gemini | throughput_rps | 6212 req/s | concurrency 64; samples 37277 | 50 |
| openai-responses → gemini | throughput_rps | 6215 req/s | concurrency 128; samples 37303 | 100 |
| openai-responses → gemini | throughput_rps | 6215 req/s | concurrency 128; samples 37303 | unbounded |
| openai → gemini | added_latency_p50_us | 336 | valid; gateway_c1_samples=16158; direct_c1_samples=241090 | — |
| openai → gemini | added_latency_p99_us | 508 | valid; gateway_c1_samples=16158; direct_c1_samples=241090 | — |
| openai → gemini | gateway_c1_p99_us | 539 | valid; gateway_c1_samples=16158 | — |
| openai → gemini | direct_c1_p99_us | 31 | valid; direct_c1_samples=241090 | — |
| openai → gemini | cpu_us_per_request | 424.58713765404383 | valid | — |
| openai → gemini | rps_per_cpu_second | 2355.229142185664 | valid | — |
| openai → gemini | cost_window_conc | 8 | valid | — |
| openai → gemini | cost_core_utilisation | 0.7070624545761913 | valid | — |
| openai → gemini | cost_window_ok | 40086 | valid | — |
| openai → gemini | cost_window_rps | 6680 | valid | — |
| openai → gemini | cost_threads | 12 | valid | — |
| openai → gemini | cost_nonvol_ctxt_per_request | 0 | valid | — |
| openai → gemini | cost_majflt | 0 | valid | — |
| openai → gemini | added_ttft_p50_us | untestable | untestable | — |
| openai → gemini | added_ttft_p99_us | untestable | untestable | — |
| openai → gemini | added_gap_p50_us | untestable | untestable | — |
| openai → gemini | added_gap_p99_us | untestable | untestable | — |
| openai → gemini | streams_sustained_fps | untestable | untestable | — |
| openai → gemini | streams_sustained | untestable | untestable | — |
| openai → gemini | idle_rss_mib | 53.0625 | valid | — |
| openai → gemini | steady_state_rss_mib | 87.9375 | valid | — |
| openai → gemini | recovered_rss_mib | 91.47265625 | valid | — |
| openai → gemini | peak_rss_mib | 91.51171875 | valid | — |
| openai → gemini | peak_rss_hwm_mib | 89.37890625 | valid | — |
| openai → gemini | time_to_plateau_s | not_measured | not_measured | — |
| openai → gemini | growth_rate_mib_per_min | 0.08107558270350812 | valid | — |
| openai → gemini | load_s | 300 | valid | — |
| openai → gemini | shape | 0 | valid | — |
| openai → gemini | idle_shape | 0 | valid | — |
| openai → gemini | idle_static | 0 | valid | — |
| openai → gemini | idle_growth_rate_mib_per_min | 0.4856319041520852 | valid | — |
| openai → gemini | throughput_rps | 4561 req/s | concurrency 2; samples 27372 | 1 |
| openai → gemini | throughput_rps | 6538 req/s | concurrency 8; samples 39236 | 5 |
| openai → gemini | throughput_rps | 6721 req/s | concurrency 16; samples 40333 | 10 |
| openai → gemini | throughput_rps | 7332 req/s | concurrency 64; samples 43996 | 50 |
| openai → gemini | throughput_rps | 7393 req/s | concurrency 128; samples 44369 | 100 |
| openai → gemini | throughput_rps | 7393 req/s | concurrency 128; samples 44369 | unbounded |
| anthropic → openai-responses | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai → openai-responses | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| openai-responses → openai-responses | added_latency_p50_us | 363 | valid; gateway_c1_samples=15015; direct_c1_samples=246146 | — |
| openai-responses → openai-responses | added_latency_p99_us | 591 | valid; gateway_c1_samples=15015; direct_c1_samples=246146 | — |
| openai-responses → openai-responses | gateway_c1_p99_us | 621 | valid; gateway_c1_samples=15015 | — |
| openai-responses → openai-responses | direct_c1_p99_us | 30 | valid; direct_c1_samples=246146 | — |
| openai-responses → openai-responses | cpu_us_per_request | 577.7154870357393 | valid | — |
| openai-responses → openai-responses | rps_per_cpu_second | 1730.9558466763708 | valid | — |
| openai-responses → openai-responses | cost_window_conc | 8 | valid | — |
| openai-responses → openai-responses | cost_core_utilisation | 0.8561883515203386 | valid | — |
| openai-responses → openai-responses | cost_window_ok | 35675 | valid | — |
| openai-responses → openai-responses | cost_window_rps | 5945 | valid | — |
| openai-responses → openai-responses | cost_threads | 12 | valid | — |
| openai-responses → openai-responses | cost_nonvol_ctxt_per_request | 0 | 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 | 52.81640625 | valid | — |
| openai-responses → openai-responses | steady_state_rss_mib | 224.859375 | valid | — |
| openai-responses → openai-responses | recovered_rss_mib | 125.69921875 | valid | — |
| openai-responses → openai-responses | peak_rss_mib | 326.2734375 | valid | — |
| openai-responses → openai-responses | peak_rss_hwm_mib | 325.96875 | valid | — |
| openai-responses → openai-responses | time_to_plateau_s | not_measured | not_measured | — |
| openai-responses → openai-responses | growth_rate_mib_per_min | 27.225418082714604 | valid | — |
| openai-responses → openai-responses | load_s | 300 | valid | — |
| openai-responses → openai-responses | shape | 1 | valid | — |
| openai-responses → openai-responses | idle_shape | not_measured | not_measured | — |
| openai-responses → openai-responses | idle_static | 1 | valid | — |
| openai-responses → openai-responses | idle_growth_rate_mib_per_min | 0.30918765034361817 | valid | — |
| openai-responses → openai-responses | throughput_rps | 2485 req/s | concurrency 1; samples 14918 | 1 |
| openai-responses → openai-responses | throughput_rps | 5918 req/s | concurrency 8; samples 35519 | 5 |
| openai-responses → openai-responses | throughput_rps | 5918 req/s | concurrency 8; samples 35519 | 10 |
| openai-responses → openai-responses | throughput_rps | 6295 req/s | concurrency 64; samples 37775 | 50 |
| openai-responses → openai-responses | throughput_rps | 6295 req/s | concurrency 64; samples 37775 | 100 |
| openai-responses → openai-responses | throughput_rps | 6295 req/s | concurrency 64; samples 37775 | unbounded |
| gemini → openai-responses | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| cohere → openai-responses | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |
| bedrock → openai-responses | — | — | not_declared: GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-preflight.json | — |

## Probe evidence and instrument provenance

Historical capability notes may reference an earlier product version. They are not a new probe or a current support claim. Full RSS samples remain in the immutable raw JSON.

| Ingress → upstream | Cell state | Probe HTTP status | Recorded reason | Recorded detail | Evidence scope | Response excerpt | Tested build |
| --- | --- | --- | --- | --- | --- | --- | --- |
| cohere → anthropic | 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai → anthropic | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai-responses → anthropic | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| anthropic → anthropic | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| anthropic → openai | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai → openai | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai-responses → cohere | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| anthropic → cohere | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai → cohere | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| anthropic → bedrock | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| bedrock → 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai-responses → bedrock | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai → bedrock | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| anthropic → gemini | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai-responses → gemini | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai → gemini | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| openai-responses → openai-responses | served | — | — | — | Recorded run evidence | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |
| 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) | — | enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05 |

**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 → anthropic | 1 | {"200":1} | 0 |
| openai → anthropic | 2 | {"200":2} | 0 |
| openai → anthropic | 4 | {"200":4} | 0 |
| openai → anthropic | 8 | {"200":8} | 0 |
| openai → anthropic | 16 | {"200":16} | 0 |
| openai → anthropic | 32 | {"200":32} | 0 |
| openai → anthropic | 64 | {"200":64} | 0 |
| openai → anthropic | 128 | {"200":128} | 0 |
| openai → anthropic | 256 | {"200":256} | 0 |
| openai → anthropic | 512 | {"200":512} | 0 |
| openai → anthropic | 1024 | {"200":1024} | 0 |
| openai → anthropic | 2048 | {"200":2048} | 0 |
| openai → anthropic | 1536 | {"200":1536} | 0 |
| openai → anthropic | 1280 | {"200":1280} | 0 |
| openai → anthropic | 1152 | {"200":1152} | 0 |
| openai → anthropic | 1088 | {"200":1088} | 0 |
| openai → anthropic | 1120 | {"200":1120} | 0 |
| openai → anthropic | 1136 | {"200":1136} | 0 |
| openai → anthropic | 1144 | {"200":1144} | 0 |
| openai → anthropic | 1140 | {"200":1140} | 0 |
| openai → anthropic | 1142 | {"200":1142} | 0 |
| openai → anthropic | 1141 | {"200":1141} | 0 |
| openai → anthropic | 1140 | {"200":1140} | 0 |
| openai → anthropic | 1140 | {"200":1140} | 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 | 1664 | {"200":1664} | 0 |
| anthropic → anthropic | 1600 | {"200":1600} | 0 |
| anthropic → anthropic | 1568 | {"200":1568} | 0 |
| anthropic → anthropic | 1552 | {"200":1552} | 0 |
| anthropic → anthropic | 1560 | {"200":1560} | 0 |
| anthropic → anthropic | 1556 | {"200":1556} | 0 |
| anthropic → anthropic | 1558 | {"200":1558} | 0 |
| anthropic → anthropic | 1559 | {"200":1559} | 0 |
| anthropic → anthropic | 1559 | {"200":1559} | 0 |
| anthropic → anthropic | 1559 | {"200":1559} | 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 | 1664 | {"200":1664} | 0 |
| anthropic → openai | 1600 | {"200":1600} | 0 |
| anthropic → openai | 1568 | {"200":1568} | 0 |
| anthropic → openai | 1552 | {"200":1552} | 0 |
| anthropic → openai | 1544 | {"200":1544} | 0 |
| anthropic → openai | 1540 | {"200":1540} | 0 |
| anthropic → openai | 1538 | {"200":1538} | 0 |
| anthropic → openai | 1537 | {"200":1537} | 0 |
| anthropic → openai | 1536 | {"200":1536} | 0 |
| anthropic → openai | 1536 | {"200":1536} | 0 |
| 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 | 1152 | {"200":1152} | 0 |
| openai → openai | 1216 | {"200":1216} | 0 |
| openai → openai | 1184 | {"200":1184} | 0 |
| openai → openai | 1200 | {"200":1200} | 0 |
| openai → openai | 1208 | {"200":1208} | 0 |
| openai → openai | 1212 | {"200":1212} | 0 |
| openai → openai | 1210 | {"200":1210} | 0 |
| openai → openai | 1209 | {"200":1209} | 0 |
| openai → openai | 1208 | {"200":1208} | 0 |
| openai → openai | 1208 | {"200":1208} | 0 |

## Recorded notes

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

- GoModel v0.2.3 source 86dbeffc254e0166367d72e0404f5a27aceeeaaa: local single-request checks using the engine's request bodies and authentication headers establish 16 exact protocol pairs; 15 are enabled on the pinned image. The 0.2.3 default configured-model allowlist remains enabled; no legacy fallback override is applied. OpenAI CompatibleProvider uses Chat Completions for chat/messages ingress and native Responses for responses ingress; the other OpenAI cross-operation cells do not select the requested upstream wire. Cohere supports translated chat/responses/messages. Native Cohere passthrough is opt-in (not in the default ENABLED_PASSTHROUGH_PROVIDERS), so cohere>cohere is not configured under the benchmark's out-of-the-box rule. Gemini and Bedrock implement upstream adapters but no native ingress/passthrough interface. These checks verify protocol routing only, not performance. Evidence: results/diagnostics/refresh-configuration/gomodel-0.2.3-correctness-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-gomodel-f6e038780b29f86549f840f902ac96740154f0f0ef2a66a3fd9dcd4bc3b83bad](https://onthebench.ai/gateways/runs/run-gomodel-f6e038780b29f86549f840f902ac96740154f0f0ef2a66a3fd9dcd4bc3b83bad)  ·  [complete raw record including RSS series](https://onthebench.ai/api/v1/runs/run-gomodel-f6e038780b29f86549f840f902ac96740154f0f0ef2a66a3fd9dcd4bc3b83bad/raw.json)

## Configuration recorded by the run

```
Not recorded.
```

## Compare a shared protocol cell

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

## Historical measurements

- [2026-11-10T17:39:52Z — enterpilot/gomodel:0.2.3@sha256:c50627e093ff88bff311e582a43d7de14e0be29d52cb333ecb3191eac7874f05](https://onthebench.ai/gateways/runs/run-gomodel-f6e038780b29f86549f840f902ac96740154f0f0ef2a66a3fd9dcd4bc3b83bad)
- [2026-08-26T19:08:45Z — enterpilot/gomodel:0.1.82](https://onthebench.ai/gateways/runs/run-gomodel-d5a944f96686f17fd771628164aa2872bb733a541b28d2b40dbbfd3876e24b19)
