{ "slug": "inference", "name": "LLM Inference Provider Benchmark", "description": "Independent benchmark of LLM inference providers all serving the same GLM 5.3 Flash model behind OpenAI-compatible serverless endpoints. One client, 600 identical structured-output requests per provider at concurrency one; end-to-end latency, time to first token, tokens per second, exact task success and failure rates. Failures stay in the denominator. No single headline metric; results are alphabetical.", "page_url": "https://openbenchmarks.com/inference", "run_id": "run-20260902-223626", "reviewed_at": "2026-09-03T03:03:46.01458+00:00", "last_updated": "2026-09-03T03:03:46.014Z", "license": "https://creativecommons.org/licenses/by/4.0/", "model": "GLM 5.3 Flash", "reasoning_effort": "low", "requests_per_provider": 600, "dataset_item_count": 721, "provider_count": 10, "metrics": [ { "key": "e2e_latency_p50", "label": "End-to-end latency, median (E2E p50)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "Client-observed time from request send to the last byte of the complete visible answer, median over completed requests." }, { "key": "e2e_latency_p95", "label": "End-to-end latency, p95", "unit": "milliseconds", "direction": "lower_is_better", "definition": "95th percentile of the complete-answer wait. Failed and timed-out requests have no completion and are excluded from percentiles but kept in failure_rate." }, { "key": "e2e_latency_p99", "label": "End-to-end latency, p99", "unit": "milliseconds", "direction": "lower_is_better", "definition": "99th percentile of the complete-answer wait; the slow-edge figure for a blocking call." }, { "key": "ttfo_p50", "label": "Time to first output token, median (TTFO p50)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "First token-bearing event, reasoning or answer." }, { "key": "ttfo_p95", "label": "Time to first output token, p95 (TTFO p95)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "95th percentile of the first token-bearing event." }, { "key": "ttfa_p50", "label": "Time to first visible answer token, median (TTFA p50)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "First visible answer token; equals TTFO when no hidden reasoning is emitted first. The figure most 'time to first token' comparisons mean." }, { "key": "ttfa_p95", "label": "Time to first visible answer token, p95 (TTFA p95)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "95th percentile of the visible-answer start." }, { "key": "ttfa_p99", "label": "Time to first visible answer token, p99 (TTFA p99)", "unit": "milliseconds", "direction": "lower_is_better", "definition": "99th percentile of the visible-answer start." }, { "key": "output_tokens_per_second_p50", "label": "Tokens per second, median", "unit": "tokens/second", "direction": "higher_is_better", "definition": "Visible answer tokens generated per second after the first answer token, counted with a pinned tiktoken o200k_base tokenizer; reasoning tokens excluded." }, { "key": "output_tokens_per_second_p95", "label": "Tokens per second, p95 floor", "unit": "tokens/second", "direction": "higher_is_better", "definition": "Speed reached by at least 95% of completed responses. A coverage floor: higher is better." }, { "key": "output_tokens_per_second_p99", "label": "Tokens per second, p99 floor", "unit": "tokens/second", "direction": "higher_is_better", "definition": "Speed reached by at least 99% of completed responses." }, { "key": "task_success_rate", "label": "Task success rate", "unit": "ratio", "direction": "higher_is_better", "definition": "Share of all submitted requests whose answer was schema-valid and exactly correct under the task-family rubric. Operational failures, malformed JSON, incomplete objects and wrong answers all count against it." }, { "key": "failure_rate", "label": "Operational failure rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Share of submitted requests that timed out, returned HTTP 429/4xx/5xx, or failed in transport. A malformed HTTP 200 is an accuracy miss, not a failure." }, { "key": "timeout_rate", "label": "Timeout rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Requests stopped by the client at the 20-second limit." }, { "key": "http_429_rate", "label": "HTTP 429 rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Rate-limited responses." }, { "key": "other_http_4xx_rate", "label": "Other HTTP 4xx rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Client-error responses other than 429." }, { "key": "http_5xx_rate", "label": "HTTP 5xx rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Server-error responses." }, { "key": "transport_error_rate", "label": "Transport error rate", "unit": "ratio", "direction": "lower_is_better", "definition": "Connection resets, TLS and stream errors with no HTTP status." }, { "key": "reasoning_emission_rate", "label": "Reasoning emission rate", "unit": "ratio", "direction": "informational", "definition": "Share of responses in which the provider emitted reasoning tokens before the visible answer; explains a gap between TTFO and TTFA." } ], "method_notes": [ "Every provider serves the same requested model under one logical request contract: one-turn JSON-object chat completion, streaming on, temperature 0, top-p 1, reasoning low, 256-token ceiling, 20-second client timeout, concurrency one.", "Latency and tokens per second are taken from the client clock and a pinned tiktoken o200k_base tokenizer, never from provider-reported usage.", "Task success counts only schema-valid, exactly correct answers; timeouts, HTTP errors and malformed responses remain in the denominator.", "Only serverless or shared endpoints are compared; dedicated capacity is outside the comparison.", "The pooled rows are balanced across the three task families (200 requests each); per-family rows are in task_families." ], "results": [ { "rank": 1, "eligible": true, "ineligibility_reasons": [], "provider": "Baseten", "provider_slug": "baseten", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.baseten.co/reference/inference-api/chat-completions", "submitted": 600, "valid_results": 590, "e2e_latency_p50": 330.760236, "e2e_latency_p95": 2285.18608, "e2e_latency_p99": 3228.816906, "ttfo_p50": 227.650189, "ttfo_p95": 498.591065, "ttfo_p99": 946.080279, "ttfa_p50": 287.05867, "ttfa_p95": 1164.803483, "ttfa_p99": 2648.757759, "output_tokens_per_second_p50": 240.4687730323997, "output_tokens_per_second_p95": 61.63003537986967, "output_tokens_per_second_p99": 22.630518982585272, "reasoning_emission_rate": 0.5733333333333334, "task_success_rate": 0.9583333333333334, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [ "operational failure rate above 1%" ], "provider": "DeepInfra", "provider_slug": "deepinfra", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.deepinfra.com/api-reference/chat-completions/openai-chat-completions", "submitted": 600, "valid_results": 576, "e2e_latency_p50": 1826.930716, "e2e_latency_p95": 11008.071913, "e2e_latency_p99": 17070.685584, "ttfo_p50": 808.306933, "ttfo_p95": 3780.499974, "ttfo_p99": 9081.563517, "ttfa_p50": 1151.697728, "ttfa_p95": 6409.773625, "ttfa_p99": 13179.837069, "output_tokens_per_second_p50": 69.43650698373987, "output_tokens_per_second_p95": 8.955625039132348, "output_tokens_per_second_p99": 4.9922234999944335, "reasoning_emission_rate": 0.6111111111111112, "task_success_rate": 0.955, "failure_rate": 0.04, "timeout_rate": 0, "http_429_rate": 0.0033333333333333335, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.03666666666666667 }, { "rank": 5, "eligible": true, "ineligibility_reasons": [], "provider": "Fireworks AI", "provider_slug": "fireworks", "model_name": "GLM 5.3 Flash", "model_key": "accounts/fireworks/models/glm-5p3-flash", "docs_url": "https://docs.fireworks.ai/api-reference/post-chatcompletions", "submitted": 600, "valid_results": 600, "e2e_latency_p50": 1954.224447, "e2e_latency_p95": 6715.757655, "e2e_latency_p99": 10487.560867, "ttfo_p50": 586.356124, "ttfo_p95": 5665.575772, "ttfo_p99": 8743.051097, "ttfa_p50": 1061.137468, "ttfa_p95": 5829.428272, "ttfa_p99": 8743.051097, "output_tokens_per_second_p50": 58.22301505501612, "output_tokens_per_second_p95": 27.853499838212947, "output_tokens_per_second_p99": 14.891372371318328, "reasoning_emission_rate": 0.5816666666666667, "task_success_rate": 0.995, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": 2, "eligible": true, "ineligibility_reasons": [], "provider": "Modal", "provider_slug": "modal", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://modal.com/docs/guide/endpoints#calling-your-endpoint", "submitted": 600, "valid_results": 600, "e2e_latency_p50": 574.117533, "e2e_latency_p95": 1898.581555, "e2e_latency_p99": 3278.852527, "ttfo_p50": 376.474939, "ttfo_p95": 812.080627, "ttfo_p99": 2721.956228, "ttfa_p50": 469.985228, "ttfa_p95": 1092.766169, "ttfa_p99": 2787.248636, "output_tokens_per_second_p50": 222.16668487091223, "output_tokens_per_second_p95": 84.71115434664893, "output_tokens_per_second_p99": 34.86570494058868, "reasoning_emission_rate": 0.5133333333333333, "task_success_rate": 0.9933333333333333, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [ "task success rate below 95%", "operational failure rate above 1%" ], "provider": "Nebius", "provider_slug": "nebius", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.tokenfactory.nebius.com/api-reference/inference/create-chat-completion", "submitted": 600, "valid_results": 511, "e2e_latency_p50": 1450.650284, "e2e_latency_p95": 7453.32705, "e2e_latency_p99": 14970.361964, "ttfo_p50": 925.170205, "ttfo_p95": 7107.37482, "ttfo_p99": 14600.476189, "ttfa_p50": 1144.277602, "ttfa_p95": 7107.37482, "ttfa_p99": 14600.476189, "output_tokens_per_second_p50": 269.2586219639911, "output_tokens_per_second_p95": 78.7936979014428, "output_tokens_per_second_p99": 39.949982976424494, "reasoning_emission_rate": 0.512720156555773, "task_success_rate": 0.8483333333333334, "failure_rate": 0.14833333333333334, "timeout_rate": 0.028333333333333332, "http_429_rate": 0, "other_http_4xx_rate": 0.115, "http_5xx_rate": 0, "transport_error_rate": 0.005 }, { "rank": null, "eligible": false, "ineligibility_reasons": [ "task success rate below 95%", "operational failure rate above 1%" ], "provider": "Novita AI", "provider_slug": "novita", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/glm-5.3-flash", "docs_url": "https://novita.ai/docs/api-reference/model-apis-llm-create-chat-completion", "submitted": 600, "valid_results": 569, "e2e_latency_p50": 1738.313375, "e2e_latency_p95": 4998.181688, "e2e_latency_p99": 7248.335321, "ttfo_p50": 1130.528242, "ttfo_p95": 1789.05836, "ttfo_p99": 4281.720074, "ttfa_p50": 1404.686841, "ttfa_p95": 3405.501919, "ttfa_p99": 5580.15286, "output_tokens_per_second_p50": 60.64171977491607, "output_tokens_per_second_p95": 29.578816841542032, "output_tokens_per_second_p99": 11.602053555859273, "reasoning_emission_rate": 0.4774305555555556, "task_success_rate": 0.9416666666666667, "failure_rate": 0.04, "timeout_rate": 0.0016666666666666668, "http_429_rate": 0.03833333333333333, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [ "operational failure rate above 1%" ], "provider": "Parasail", "provider_slug": "parasail", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.parasail.io/parasail-docs/api-reference/chat-completions", "submitted": 600, "valid_results": 592, "e2e_latency_p50": 1590.817551, "e2e_latency_p95": 7740.687499, "e2e_latency_p99": 11512.25082, "ttfo_p50": 856.198122, "ttfo_p95": 4956.699523, "ttfo_p99": 10905.492396, "ttfa_p50": 1108.915037, "ttfa_p95": 5539.380056, "ttfa_p99": 10905.492396, "output_tokens_per_second_p50": 52.350789822473565, "output_tokens_per_second_p95": 22.81741815548379, "output_tokens_per_second_p99": 15.595400478168234, "reasoning_emission_rate": 0.527027027027027, "task_success_rate": 0.9833333333333333, "failure_rate": 0.013333333333333334, "timeout_rate": 0.008333333333333333, "http_429_rate": 0.0016666666666666668, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.0033333333333333335 }, { "rank": 3, "eligible": true, "ineligibility_reasons": [], "provider": "Telnyx", "provider_slug": "telnyx", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://developers.telnyx.com/api-reference/openai-chat/create-a-chat-completion-openai-compatible", "submitted": 600, "valid_results": 600, "e2e_latency_p50": 669.374216, "e2e_latency_p95": 2331.560924, "e2e_latency_p99": 3478.362819, "ttfo_p50": 496.642397, "ttfo_p95": 801.519196, "ttfo_p99": 2895.539492, "ttfa_p50": 573.537664, "ttfa_p95": 1343.680012, "ttfa_p99": 2995.55873, "output_tokens_per_second_p50": 184.29680893335956, "output_tokens_per_second_p95": 67.48298309733144, "output_tokens_per_second_p99": 47.27662619543005, "reasoning_emission_rate": 0.5283333333333333, "task_success_rate": 0.995, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": 6, "eligible": true, "ineligibility_reasons": [], "provider": "Together AI", "provider_slug": "together", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.together.ai/docs/inference/chat/overview", "submitted": 600, "valid_results": 594, "e2e_latency_p50": 608.640461, "e2e_latency_p95": 3662.672044, "e2e_latency_p99": 12449.663781, "ttfo_p50": 377.296148, "ttfo_p95": 2228.134735, "ttfo_p99": 12197.729187, "ttfa_p50": 393.151152, "ttfa_p95": 2264.080438, "ttfa_p99": 12197.729187, "output_tokens_per_second_p50": 151.6462247555697, "output_tokens_per_second_p95": 34.022559958798524, "output_tokens_per_second_p99": 21.710334625675415, "reasoning_emission_rate": 0.5286195286195287, "task_success_rate": 0.985, "failure_rate": 0.01, "timeout_rate": 0.0033333333333333335, "http_429_rate": 0.005, "other_http_4xx_rate": 0, "http_5xx_rate": 0.0016666666666666668, "transport_error_rate": 0 }, { "rank": 4, "eligible": true, "ineligibility_reasons": [], "provider": "Z.AI", "provider_slug": "z-ai", "model_name": "GLM 5.3 Flash", "model_key": "glm-5.3-flash", "docs_url": "https://docs.z.ai/api-reference/llm/chat-completion", "submitted": 600, "valid_results": 591, "e2e_latency_p50": 1748.762339, "e2e_latency_p95": 4826.123173, "e2e_latency_p99": 8591.183064, "ttfo_p50": 1047.263091, "ttfo_p95": 1888.236768, "ttfo_p99": 4755.838108, "ttfa_p50": 1391.248712, "ttfa_p95": 3137.421355, "ttfa_p99": 4997.114816, "output_tokens_per_second_p50": 56.273675958754964, "output_tokens_per_second_p95": 27.84024039317547, "output_tokens_per_second_p99": 11.495453450038609, "reasoning_emission_rate": 0.5258764607679466, "task_success_rate": 0.975, "failure_rate": 0.0016666666666666668, "timeout_rate": 0.0016666666666666668, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 } ], "task_families": { "contract-terms-extraction": [ { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Baseten", "provider_slug": "baseten", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.baseten.co/reference/inference-api/chat-completions", "submitted": 200, "valid_results": 190, "e2e_latency_p50": 1565.795885, "e2e_latency_p95": 2756.115123, "e2e_latency_p99": 4553.515538, "ttfo_p50": 343.101436, "ttfo_p95": 574.560136, "ttfo_p99": 1060.035597, "ttfa_p50": 602.73408, "ttfa_p95": 2071.83179, "ttfa_p99": 3043.466142, "output_tokens_per_second_p50": 93.05789558027074, "output_tokens_per_second_p95": 59.431946251530775, "output_tokens_per_second_p99": 36.652789705149154, "reasoning_emission_rate": 0.855, "task_success_rate": 0.9, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "DeepInfra", "provider_slug": "deepinfra", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.deepinfra.com/api-reference/chat-completions/openai-chat-completions", "submitted": 200, "valid_results": 184, "e2e_latency_p50": 3514.810086, "e2e_latency_p95": 13173.069092, "e2e_latency_p99": 19842.254133, "ttfo_p50": 906.615967, "ttfo_p95": 2328.064199, "ttfo_p99": 4623.357453, "ttfa_p50": 1910.765512, "ttfa_p95": 6460.26541, "ttfa_p99": 12408.329269, "output_tokens_per_second_p50": 58.56626855059653, "output_tokens_per_second_p95": 11.81846417070849, "output_tokens_per_second_p99": 5.343425958638145, "reasoning_emission_rate": 0.8641304347826086, "task_success_rate": 0.92, "failure_rate": 0.08, "timeout_rate": 0, "http_429_rate": 0.005, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.075 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Fireworks AI", "provider_slug": "fireworks", "model_name": "GLM 5.3 Flash", "model_key": "accounts/fireworks/models/glm-5p3-flash", "docs_url": "https://docs.fireworks.ai/api-reference/post-chatcompletions", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 4206.740776, "e2e_latency_p95": 8190.485067, "e2e_latency_p99": 11907.404015, "ttfo_p50": 601.709071, "ttfo_p95": 6798.943842, "ttfo_p99": 10877.118861, "ttfa_p50": 2023.946045, "ttfa_p95": 6911.608112, "ttfa_p99": 11231.419264, "output_tokens_per_second_p50": 37.599169969264814, "output_tokens_per_second_p95": 25.74890193662854, "output_tokens_per_second_p99": 14.857892676026324, "reasoning_emission_rate": 0.915, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Modal", "provider_slug": "modal", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://modal.com/docs/guide/endpoints#calling-your-endpoint", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 1301.518133, "e2e_latency_p95": 2378.377627, "e2e_latency_p99": 3374.587828, "ttfo_p50": 388.915481, "ttfo_p95": 814.590898, "ttfo_p99": 2993.253979, "ttfa_p50": 659.50789, "ttfa_p95": 1436.803478, "ttfa_p99": 3004.185424, "output_tokens_per_second_p50": 129.88636524407642, "output_tokens_per_second_p95": 69.02599633173595, "output_tokens_per_second_p99": 31.832248215562053, "reasoning_emission_rate": 0.825, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Nebius", "provider_slug": "nebius", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.tokenfactory.nebius.com/api-reference/inference/create-chat-completion", "submitted": 200, "valid_results": 171, "e2e_latency_p50": 2253.915377, "e2e_latency_p95": 7335.944696, "e2e_latency_p99": 13796.913638, "ttfo_p50": 954.464447, "ttfo_p95": 6920.752403, "ttfo_p99": 13014.037433, "ttfa_p50": 1494.647724, "ttfa_p95": 6993.893283, "ttfa_p99": 13387.4183, "output_tokens_per_second_p50": 111.94900682437505, "output_tokens_per_second_p95": 75.53102851904211, "output_tokens_per_second_p99": 51.71697362549094, "reasoning_emission_rate": 0.8070175438596491, "task_success_rate": 0.855, "failure_rate": 0.145, "timeout_rate": 0.025, "http_429_rate": 0, "other_http_4xx_rate": 0.115, "http_5xx_rate": 0, "transport_error_rate": 0.005 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Novita AI", "provider_slug": "novita", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/glm-5.3-flash", "docs_url": "https://novita.ai/docs/api-reference/model-apis-llm-create-chat-completion", "submitted": 200, "valid_results": 187, "e2e_latency_p50": 4073.120351, "e2e_latency_p95": 6445.719442, "e2e_latency_p99": 8987.518772, "ttfo_p50": 1147.423902, "ttfo_p95": 1914.169819, "ttfo_p99": 5580.15286, "ttfa_p50": 2128.691459, "ttfa_p95": 4127.350823, "ttfa_p99": 6408.551975, "output_tokens_per_second_p50": 42.33288846902886, "output_tokens_per_second_p95": 26.837614839172886, "output_tokens_per_second_p99": 11.602053555859273, "reasoning_emission_rate": 0.7371134020618557, "task_success_rate": 0.93, "failure_rate": 0.03, "timeout_rate": 0, "http_429_rate": 0.03, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Parasail", "provider_slug": "parasail", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.parasail.io/parasail-docs/api-reference/chat-completions", "submitted": 200, "valid_results": 195, "e2e_latency_p50": 3597.96159, "e2e_latency_p95": 8619.931653, "e2e_latency_p99": 12176.508294, "ttfo_p50": 826.421047, "ttfo_p95": 4763.614353, "ttfo_p99": 8402.88483, "ttfa_p50": 1563.291304, "ttfa_p95": 5792.919777, "ttfa_p99": 10189.425568, "output_tokens_per_second_p50": 42.273671135628675, "output_tokens_per_second_p95": 24.197661620969967, "output_tokens_per_second_p99": 13.057803477461773, "reasoning_emission_rate": 0.8358974358974359, "task_success_rate": 0.975, "failure_rate": 0.025, "timeout_rate": 0.01, "http_429_rate": 0.005, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.01 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Telnyx", "provider_slug": "telnyx", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://developers.telnyx.com/api-reference/openai-chat/create-a-chat-completion-openai-compatible", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 1700.708948, "e2e_latency_p95": 2505.710752, "e2e_latency_p99": 3478.362819, "ttfo_p50": 492.114852, "ttfo_p95": 735.082855, "ttfo_p99": 1620.916843, "ttfa_p50": 778.789527, "ttfa_p95": 1366.305977, "ttfa_p99": 1870.252215, "output_tokens_per_second_p50": 91.38032870059284, "output_tokens_per_second_p95": 62.39403586766855, "output_tokens_per_second_p99": 38.351240548667015, "reasoning_emission_rate": 0.825, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Together AI", "provider_slug": "together", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.together.ai/docs/inference/chat/overview", "submitted": 200, "valid_results": 197, "e2e_latency_p50": 1531.713283, "e2e_latency_p95": 2813.71193, "e2e_latency_p99": 6128.717204, "ttfo_p50": 510.491511, "ttfo_p95": 1267.621612, "ttfo_p99": 4212.99712, "ttfa_p50": 516.20085, "ttfa_p95": 1324.92443, "ttfa_p99": 4246.078948, "output_tokens_per_second_p50": 86.83010616508689, "output_tokens_per_second_p95": 51.585738841355024, "output_tokens_per_second_p99": 23.222886984877352, "reasoning_emission_rate": 0.751269035532995, "task_success_rate": 0.985, "failure_rate": 0.015, "timeout_rate": 0.005, "http_429_rate": 0.01, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Z.AI", "provider_slug": "z-ai", "model_name": "GLM 5.3 Flash", "model_key": "glm-5.3-flash", "docs_url": "https://docs.z.ai/api-reference/llm/chat-completion", "submitted": 200, "valid_results": 192, "e2e_latency_p50": 4070.542359, "e2e_latency_p95": 5286.259498, "e2e_latency_p99": 8736.303372, "ttfo_p50": 1053.149208, "ttfo_p95": 1905.910082, "ttfo_p99": 2578.283054, "ttfa_p50": 2048.588088, "ttfa_p95": 3539.726651, "ttfa_p99": 4543.1483, "output_tokens_per_second_p50": 41.47145052252333, "output_tokens_per_second_p95": 26.533330326877707, "output_tokens_per_second_p99": 11.863456892775167, "reasoning_emission_rate": 0.725, "task_success_rate": 0.96, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 } ], "meeting-notes-lookup": [ { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Baseten", "provider_slug": "baseten", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.baseten.co/reference/inference-api/chat-completions", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 239.19674, "e2e_latency_p95": 599.053569, "e2e_latency_p99": 1432.149556, "ttfo_p50": 198.088607, "ttfo_p95": 396.307962, "ttfo_p99": 946.080279, "ttfa_p50": 211.691263, "ttfa_p95": 453.921739, "ttfa_p99": 946.080279, "output_tokens_per_second_p50": 314.1170812785278, "output_tokens_per_second_p95": 124.59420912031611, "output_tokens_per_second_p99": 14.90256553626733, "reasoning_emission_rate": 0.215, "task_success_rate": 0.995, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "DeepInfra", "provider_slug": "deepinfra", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.deepinfra.com/api-reference/chat-completions/openai-chat-completions", "submitted": 200, "valid_results": 196, "e2e_latency_p50": 1014.948359, "e2e_latency_p95": 7035.474438, "e2e_latency_p99": 15694.913265, "ttfo_p50": 818.183503, "ttfo_p95": 6210.621437, "ttfo_p99": 14879.582956, "ttfa_p50": 861.186612, "ttfa_p95": 6210.621437, "ttfa_p99": 14879.582956, "output_tokens_per_second_p50": 78.4615431002539, "output_tokens_per_second_p95": 8.233568674044262, "output_tokens_per_second_p99": 4.9922234999944335, "reasoning_emission_rate": 0.12244897959183673, "task_success_rate": 0.98, "failure_rate": 0.02, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.02 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Fireworks AI", "provider_slug": "fireworks", "model_name": "GLM 5.3 Flash", "model_key": "accounts/fireworks/models/glm-5p3-flash", "docs_url": "https://docs.fireworks.ai/api-reference/post-chatcompletions", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 734.210747, "e2e_latency_p95": 5358.253125, "e2e_latency_p99": 7542.75799, "ttfo_p50": 568.203256, "ttfo_p95": 5281.302325, "ttfo_p99": 7468.241637, "ttfa_p50": 574.164295, "ttfa_p95": 5281.302325, "ttfa_p99": 7468.241637, "output_tokens_per_second_p50": 76.21266087124344, "output_tokens_per_second_p95": 33.496894682341654, "output_tokens_per_second_p99": 30.976962598126235, "reasoning_emission_rate": 0.09, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Modal", "provider_slug": "modal", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://modal.com/docs/guide/endpoints#calling-your-endpoint", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 412.652983, "e2e_latency_p95": 775.596503, "e2e_latency_p99": 1627.231234, "ttfo_p50": 378.467927, "ttfo_p95": 748.39682, "ttfo_p99": 1612.210754, "ttfa_p50": 378.741428, "ttfa_p95": 760.530326, "ttfa_p99": 1612.210754, "output_tokens_per_second_p50": 289.67006744843405, "output_tokens_per_second_p95": 159.4186422980606, "output_tokens_per_second_p99": 33.58094613355316, "reasoning_emission_rate": 0.045, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Nebius", "provider_slug": "nebius", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.tokenfactory.nebius.com/api-reference/inference/create-chat-completion", "submitted": 200, "valid_results": 169, "e2e_latency_p50": 948.693153, "e2e_latency_p95": 8017.100552, "e2e_latency_p99": 18626.030371, "ttfo_p50": 923.361873, "ttfo_p95": 8003.250437, "ttfo_p99": 18605.801739, "ttfa_p50": 923.821501, "ttfa_p95": 8003.250437, "ttfa_p99": 18605.801739, "output_tokens_per_second_p50": 385.24201673980485, "output_tokens_per_second_p95": 94.93067655700868, "output_tokens_per_second_p99": 9.459813211731218, "reasoning_emission_rate": 0.047337278106508875, "task_success_rate": 0.845, "failure_rate": 0.155, "timeout_rate": 0.04, "http_429_rate": 0, "other_http_4xx_rate": 0.11, "http_5xx_rate": 0, "transport_error_rate": 0.005 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Novita AI", "provider_slug": "novita", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/glm-5.3-flash", "docs_url": "https://novita.ai/docs/api-reference/model-apis-llm-create-chat-completion", "submitted": 200, "valid_results": 189, "e2e_latency_p50": 1256.000625, "e2e_latency_p95": 1930.330898, "e2e_latency_p99": 6131.607916, "ttfo_p50": 1102.736173, "ttfo_p95": 1745.527779, "ttfo_p99": 6027.226714, "ttfa_p50": 1126.632809, "ttfa_p95": 1811.076321, "ttfa_p99": 6027.226714, "output_tokens_per_second_p50": 81.15661749314174, "output_tokens_per_second_p95": 36.469908793742334, "output_tokens_per_second_p99": 29.46314756867895, "reasoning_emission_rate": 0.10052910052910052, "task_success_rate": 0.945, "failure_rate": 0.055, "timeout_rate": 0, "http_429_rate": 0.055, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Parasail", "provider_slug": "parasail", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.parasail.io/parasail-docs/api-reference/chat-completions", "submitted": 200, "valid_results": 198, "e2e_latency_p50": 1081.450258, "e2e_latency_p95": 5433.896234, "e2e_latency_p99": 11064.512807, "ttfo_p50": 893.785486, "ttfo_p95": 5043.716152, "ttfo_p99": 10905.492396, "ttfa_p50": 895.59822, "ttfa_p95": 5185.285203, "ttfa_p99": 10905.492396, "output_tokens_per_second_p50": 74.7885211220608, "output_tokens_per_second_p95": 22.71574209455882, "output_tokens_per_second_p99": 16.232171978634366, "reasoning_emission_rate": 0.0707070707070707, "task_success_rate": 0.99, "failure_rate": 0.01, "timeout_rate": 0.01, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Telnyx", "provider_slug": "telnyx", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://developers.telnyx.com/api-reference/openai-chat/create-a-chat-completion-openai-compatible", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 526.573056, "e2e_latency_p95": 747.822867, "e2e_latency_p99": 2922.535946, "ttfo_p50": 488.522785, "ttfo_p95": 646.140422, "ttfo_p99": 2895.539492, "ttfa_p50": 492.011056, "ttfa_p95": 693.298217, "ttfa_p99": 2895.539492, "output_tokens_per_second_p50": 339.8091646833767, "output_tokens_per_second_p95": 132.27628229620117, "output_tokens_per_second_p99": 89.84953616973944, "reasoning_emission_rate": 0.07, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Together AI", "provider_slug": "together", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.together.ai/docs/inference/chat/overview", "submitted": 200, "valid_results": 197, "e2e_latency_p50": 385.499745, "e2e_latency_p95": 12337.107385, "e2e_latency_p99": 12842.647382, "ttfo_p50": 333.86177, "ttfo_p95": 12042.763673, "ttfo_p99": 12805.205776, "ttfa_p50": 334.184952, "ttfa_p95": 12042.763673, "ttfa_p99": 12805.205776, "output_tokens_per_second_p50": 233.95792734592536, "output_tokens_per_second_p95": 27.24917596757835, "output_tokens_per_second_p99": 21.273776742945667, "reasoning_emission_rate": 0.06091370558375635, "task_success_rate": 0.985, "failure_rate": 0.015, "timeout_rate": 0.005, "http_429_rate": 0.005, "other_http_4xx_rate": 0, "http_5xx_rate": 0.005, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Z.AI", "provider_slug": "z-ai", "model_name": "GLM 5.3 Flash", "model_key": "glm-5.3-flash", "docs_url": "https://docs.z.ai/api-reference/llm/chat-completion", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 1205.25738, "e2e_latency_p95": 2544.603754, "e2e_latency_p99": 7571.203048, "ttfo_p50": 1030.56902, "ttfo_p95": 1958.108035, "ttfo_p99": 6940.337216, "ttfa_p50": 1059.881944, "ttfa_p95": 1963.101894, "ttfa_p99": 7371.190293, "output_tokens_per_second_p50": 77.28662793633372, "output_tokens_per_second_p95": 37.90184609137665, "output_tokens_per_second_p99": 10.864904456811622, "reasoning_emission_rate": 0.1, "task_success_rate": 1, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 } ], "ticket-triage": [ { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Baseten", "provider_slug": "baseten", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.baseten.co/reference/inference-api/chat-completions", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 273.969101, "e2e_latency_p95": 749.097836, "e2e_latency_p99": 1200.842341, "ttfo_p50": 184.78025, "ttfo_p95": 415.202421, "ttfo_p99": 512.835646, "ttfa_p50": 230.599744, "ttfa_p95": 537.636511, "ttfa_p99": 1125.669746, "output_tokens_per_second_p50": 313.0466728027025, "output_tokens_per_second_p95": 49.10431222624889, "output_tokens_per_second_p99": 23.067488030712976, "reasoning_emission_rate": 0.65, "task_success_rate": 0.98, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "DeepInfra", "provider_slug": "deepinfra", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.deepinfra.com/api-reference/chat-completions/openai-chat-completions", "submitted": 200, "valid_results": 196, "e2e_latency_p50": 1225.660592, "e2e_latency_p95": 8425.432842, "e2e_latency_p99": 17070.685584, "ttfo_p50": 729.765889, "ttfo_p95": 2815.488522, "ttfo_p99": 11006.900586, "ttfa_p50": 981.572758, "ttfa_p95": 6230.924397, "ttfa_p99": 15379.825153, "output_tokens_per_second_p50": 89.53681642343423, "output_tokens_per_second_p95": 7.704862807283977, "output_tokens_per_second_p99": 2.873416572998016, "reasoning_emission_rate": 0.8622448979591837, "task_success_rate": 0.965, "failure_rate": 0.02, "timeout_rate": 0, "http_429_rate": 0.005, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0.015 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Fireworks AI", "provider_slug": "fireworks", "model_name": "GLM 5.3 Flash", "model_key": "accounts/fireworks/models/glm-5p3-flash", "docs_url": "https://docs.fireworks.ai/api-reference/post-chatcompletions", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 1072.620503, "e2e_latency_p95": 5706.343636, "e2e_latency_p99": 7698.420012, "ttfo_p50": 597.042289, "ttfo_p95": 5618.146716, "ttfo_p99": 7462.789288, "ttfa_p50": 836.120879, "ttfa_p95": 5618.146716, "ttfa_p99": 7541.71613, "output_tokens_per_second_p50": 63.340597021864475, "output_tokens_per_second_p95": 29.585571773900725, "output_tokens_per_second_p99": 10.903614278310746, "reasoning_emission_rate": 0.74, "task_success_rate": 0.985, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Modal", "provider_slug": "modal", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://modal.com/docs/guide/endpoints#calling-your-endpoint", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 492.948614, "e2e_latency_p95": 1207.429462, "e2e_latency_p99": 2756.256493, "ttfo_p50": 362.026429, "ttfo_p95": 853.669706, "ttfo_p99": 2650.482911, "ttfa_p50": 446.49436, "ttfa_p95": 1056.454708, "ttfa_p99": 2715.746952, "output_tokens_per_second_p50": 270.32068142437345, "output_tokens_per_second_p95": 99.43251543350074, "output_tokens_per_second_p99": 43.55028325503442, "reasoning_emission_rate": 0.67, "task_success_rate": 0.98, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Nebius", "provider_slug": "nebius", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.tokenfactory.nebius.com/api-reference/inference/create-chat-completion", "submitted": 200, "valid_results": 171, "e2e_latency_p50": 1019.112608, "e2e_latency_p95": 6191.692893, "e2e_latency_p99": 14989.353481, "ttfo_p50": 901.839286, "ttfo_p95": 6024.892688, "ttfo_p99": 14932.776144, "ttfa_p50": 968.560304, "ttfa_p95": 6099.262483, "ttfa_p99": 14932.776144, "output_tokens_per_second_p50": 323.29510997287275, "output_tokens_per_second_p95": 122.42840162620827, "output_tokens_per_second_p99": 45.503428706104714, "reasoning_emission_rate": 0.6783625730994152, "task_success_rate": 0.845, "failure_rate": 0.145, "timeout_rate": 0.02, "http_429_rate": 0, "other_http_4xx_rate": 0.12, "http_5xx_rate": 0, "transport_error_rate": 0.005 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Novita AI", "provider_slug": "novita", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/glm-5.3-flash", "docs_url": "https://novita.ai/docs/api-reference/model-apis-llm-create-chat-completion", "submitted": 200, "valid_results": 193, "e2e_latency_p50": 1631.967778, "e2e_latency_p95": 2771.97776, "e2e_latency_p99": 4105.745393, "ttfo_p50": 1149.765967, "ttfo_p95": 1657.87037, "ttfo_p99": 2849.310113, "ttfa_p50": 1398.721678, "ttfa_p95": 2636.410145, "ttfa_p99": 3353.594192, "output_tokens_per_second_p50": 65.42929555862065, "output_tokens_per_second_p95": 25.861306781529667, "output_tokens_per_second_p99": 9.042100081273109, "reasoning_emission_rate": 0.5854922279792746, "task_success_rate": 0.95, "failure_rate": 0.035, "timeout_rate": 0.005, "http_429_rate": 0.03, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Parasail", "provider_slug": "parasail", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.parasail.io/parasail-docs/api-reference/chat-completions", "submitted": 200, "valid_results": 199, "e2e_latency_p50": 1475.0314, "e2e_latency_p95": 6081.691908, "e2e_latency_p99": 13124.962056, "ttfo_p50": 856.581062, "ttfo_p95": 4987.45965, "ttfo_p99": 12138.631834, "ttfa_p50": 1146.621983, "ttfa_p95": 5667.810873, "ttfa_p99": 12816.731507, "output_tokens_per_second_p50": 48.81591258761337, "output_tokens_per_second_p95": 22.13422608979785, "output_tokens_per_second_p99": 11.01573730630244, "reasoning_emission_rate": 0.678391959798995, "task_success_rate": 0.985, "failure_rate": 0.005, "timeout_rate": 0.005, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Telnyx", "provider_slug": "telnyx", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://developers.telnyx.com/api-reference/openai-chat/create-a-chat-completion-openai-compatible", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 644.890314, "e2e_latency_p95": 2241.323001, "e2e_latency_p99": 3782.646858, "ttfo_p50": 510.711824, "ttfo_p95": 2205.695243, "ttfo_p99": 3733.215384, "ttfa_p50": 583.246982, "ttfa_p95": 2208.702355, "ttfa_p99": 3757.514286, "output_tokens_per_second_p50": 211.856551928689, "output_tokens_per_second_p95": 93.24563414078429, "output_tokens_per_second_p99": 34.324582167647876, "reasoning_emission_rate": 0.69, "task_success_rate": 0.985, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Together AI", "provider_slug": "together", "model_name": "GLM 5.3 Flash", "model_key": "zai-org/GLM-5.3-Flash", "docs_url": "https://docs.together.ai/docs/inference/chat/overview", "submitted": 200, "valid_results": 200, "e2e_latency_p50": 417.437267, "e2e_latency_p95": 1479.520259, "e2e_latency_p99": 6829.374523, "ttfo_p50": 330.970952, "ttfo_p95": 1155.48837, "ttfo_p99": 6759.589061, "ttfa_p50": 347.826619, "ttfa_p95": 1314.804496, "ttfa_p99": 6759.589061, "output_tokens_per_second_p50": 178.56771372524116, "output_tokens_per_second_p95": 42.07862000921487, "output_tokens_per_second_p99": 21.173498302363605, "reasoning_emission_rate": 0.77, "task_success_rate": 0.985, "failure_rate": 0, "timeout_rate": 0, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 }, { "rank": null, "eligible": false, "ineligibility_reasons": [], "provider": "Z.AI", "provider_slug": "z-ai", "model_name": "GLM 5.3 Flash", "model_key": "glm-5.3-flash", "docs_url": "https://docs.z.ai/api-reference/llm/chat-completion", "submitted": 200, "valid_results": 199, "e2e_latency_p50": 1619.24591, "e2e_latency_p95": 3319.047528, "e2e_latency_p99": 5751.445199, "ttfo_p50": 1060.325921, "ttfo_p95": 1835.732335, "ttfo_p99": 4060.879717, "ttfa_p50": 1394.267281, "ttfa_p95": 2646.299299, "ttfa_p99": 4533.572406, "output_tokens_per_second_p50": 57.246806775319925, "output_tokens_per_second_p95": 21.530071799452287, "output_tokens_per_second_p99": 9.202832150734634, "reasoning_emission_rate": 0.7537688442211056, "task_success_rate": 0.965, "failure_rate": 0.005, "timeout_rate": 0.005, "http_429_rate": 0, "other_http_4xx_rate": 0, "http_5xx_rate": 0, "transport_error_rate": 0 } ] }, "intent_pages": [ { "title": "Fastest LLM inference provider: time to first token, tokens per second and end-to-end latency", "url": "https://openbenchmarks.com/inference/fastest-inference-provider" }, { "title": "Fastest LLM API provider: time to first token and complete-response latency", "url": "https://openbenchmarks.com/inference/fastest-llm-api-provider" }, { "title": "Fastest LLM inference API by tokens per second", "url": "https://openbenchmarks.com/inference/fastest-streaming-inference-api" }, { "title": "Same model, different accuracy: LLM inference provider endpoint accuracy", "url": "https://openbenchmarks.com/inference/most-accurate-inference-provider" }, { "title": "Most reliable LLM API provider: failure rate, errors, timeouts and usable answers", "url": "https://openbenchmarks.com/inference/most-accurate-llm-api-provider" }, { "title": "Best LLM inference provider: speed, accuracy and failures compared", "url": "https://openbenchmarks.com/inference/best-inference-provider" }, { "title": "Lowest-latency LLM inference provider: p99 tail analysis", "url": "https://openbenchmarks.com/inference/lowest-latency-llm-inference-provider" }, { "title": "Best LLM inference provider for AI agents: tool calling latency and accuracy", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-my-ai-agent" }, { "title": "Fastest LLM inference provider for AI agents", "url": "https://openbenchmarks.com/inference/fastest-inference-provider-for-my-ai-agent" }, { "title": "Most accurate LLM inference provider for AI agents", "url": "https://openbenchmarks.com/inference/most-accurate-inference-provider-for-my-ai-agent" }, { "title": "GLM 5.3 Flash API providers compared: latency, tokens per second and accuracy", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-glm-5-3-flash" }, { "title": "Fastest GLM 5.3 Flash API provider", "url": "https://openbenchmarks.com/inference/fastest-inference-provider-for-glm-5-3-flash" }, { "title": "Most accurate GLM 5.3 Flash provider", "url": "https://openbenchmarks.com/inference/most-accurate-inference-provider-for-glm-5-3-flash" }, { "title": "Best LLM inference provider for small tasks and short prompts", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-small-tasks" }, { "title": "Fastest LLM inference provider for small tasks", "url": "https://openbenchmarks.com/inference/fastest-inference-provider-for-small-tasks" }, { "title": "Best LLM inference provider for chat applications", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-chat-applications" }, { "title": "Best LLM API for text classification and ticket routing", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-classification" }, { "title": "Fastest LLM API for classification and ticket routing", "url": "https://openbenchmarks.com/inference/fastest-inference-provider-for-classification" }, { "title": "Most reliable LLM API for structured output and JSON schema extraction", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-structured-extraction" }, { "title": "Best LLM inference provider for RAG answer extraction", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-rag" }, { "title": "Best LLM API for customer support automation", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-customer-support-automation" }, { "title": "Best LLM inference API for chatbots and copilots: time to first token and tokens per second", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-chatbots-and-copilots" }, { "title": "Lowest-latency LLM API for synchronous workflows with an SLA", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-synchronous-workflows" }, { "title": "Best LLM inference provider for enterprise AI programs", "url": "https://openbenchmarks.com/inference/best-inference-provider-for-ai-transformation" }, { "title": "Baseten alternatives, benchmarked", "url": "https://openbenchmarks.com/inference/baseten-alternatives" }, { "title": "Together AI alternatives, benchmarked", "url": "https://openbenchmarks.com/inference/together-ai-alternatives" }, { "title": "Fireworks AI alternatives, benchmarked", "url": "https://openbenchmarks.com/inference/fireworks-ai-alternatives" }, { "title": "Baseten vs Fireworks AI", "url": "https://openbenchmarks.com/inference/baseten-vs-fireworks-ai" }, { "title": "Baseten vs Together AI", "url": "https://openbenchmarks.com/inference/baseten-vs-together-ai" }, { "title": "Fireworks AI vs Together AI", "url": "https://openbenchmarks.com/inference/fireworks-ai-vs-together-ai" }, { "title": "DeepInfra vs Together AI", "url": "https://openbenchmarks.com/inference/deepinfra-vs-together-ai" } ], "methodology_url": "https://openbenchmarks.com/inference#methodology" }