{ "base_url": "http://127.0.0.1:39279/v1", "completed": 100, "concurrency": 8, "context_distribution": { "4096": 100 }, "context_seed": 0, "context_tokens": 4096, "engine": "vllm-0.27.1", "failed": 0, "failures": [], "finished_at": "2026-09-05T05:07:17Z", "gpu": "Primary Node / RTX PRO 6000 Blackwell Max-Q / TP=1", "manifest": null, "max_context_tokens": 262144, "max_tokens": 512, "measurement_protocol": "capacity-v3", "metrics": { "completion_tokens_max": 512.0, "completion_tokens_mean": 509.29, "completion_tokens_mean_ci95_high": 513.4202858480089, "completion_tokens_mean_ci95_low": 505.1597141519911, "completion_tokens_min": 310.0, "completion_tokens_p25": 512.0, "completion_tokens_p50": 512.0, "completion_tokens_p75": 512.0, "completion_tokens_p90": 512.0, "completion_tokens_p95": 512.0, "completion_tokens_p99": 512.0, "completion_tokens_samples": 100, "completion_tokens_stddev": 21.07288697963726, "content_chunks": 17392, "content_chunks_s": null, "decode_tok_s_max": 118.71561966206109, "decode_tok_s_mean": 75.43114811939104, "decode_tok_s_mean_ci95_high": 77.73869096107548, "decode_tok_s_mean_ci95_low": 73.1236052777066, "decode_tok_s_min": 44.98881358879211, "decode_tok_s_p25": 68.63278324792834, "decode_tok_s_p50": 72.87895534373254, "decode_tok_s_p75": 82.15547410583004, "decode_tok_s_p90": 92.25439834586915, "decode_tok_s_p95": 97.25613990825097, "decode_tok_s_p99": 107.54471645148837, "decode_tok_s_samples": 100, "decode_tok_s_stddev": 11.773177763696141, "e2e_max_ms": 11837.91989996098, "e2e_mean_ci95_high_ms": 8120.892984751331, "e2e_mean_ci95_low_ms": 7724.242541263161, "e2e_mean_ms": 7922.567763007246, "e2e_min_ms": 4468.927499954589, "e2e_p25_ms": 7508.464200072922, "e2e_p50_ms": 7858.428900013678, "e2e_p75_ms": 8219.337899936363, "e2e_p90_ms": 8763.958800002001, "e2e_p95_ms": 9680.693799979053, "e2e_p99_ms": 10750.324600026943, "e2e_samples": 100, "e2e_stddev_ms": 1011.8633762453301, "effective_prefill_tok_s_max": 7784.190041522781, "effective_prefill_tok_s_mean": 5324.782017336941, "effective_prefill_tok_s_mean_ci95_high": 5829.362503813258, "effective_prefill_tok_s_mean_ci95_low": 4820.201530860624, "effective_prefill_tok_s_min": 1048.9894585315087, "effective_prefill_tok_s_p25": 2839.7010822804364, "effective_prefill_tok_s_p50": 7595.325069650224, "effective_prefill_tok_s_p75": 7662.347430923353, "effective_prefill_tok_s_p90": 7685.261507011951, "effective_prefill_tok_s_p95": 7703.330813253954, "effective_prefill_tok_s_p99": 7746.756898260891, "effective_prefill_tok_s_samples": 100, "effective_prefill_tok_s_stddev": 2574.390237124067, "generation_max_ms": 11358.379099983722, "generation_mean_ci95_high_ms": 7115.836012033049, "generation_mean_ci95_low_ms": 6684.025675977811, "generation_mean_ms": 6899.93084400543, "generation_min_ms": 3177.1772999782115, "generation_p25_ms": 6217.253300012089, "generation_p50_ms": 7006.260500056669, "generation_p75_ms": 7442.799299955368, "generation_p90_ms": 7825.867300038226, "generation_p95_ms": 8280.512199969962, "generation_p99_ms": 9483.01960004028, "generation_samples": 100, "generation_stddev_ms": 1101.5569797327514, "mean_inter_token_latency_ms_max": 22.227747749478908, "mean_inter_token_latency_ms_mean": 13.561711619536988, "mean_inter_token_latency_ms_mean_ci95_high": 13.962753643269135, "mean_inter_token_latency_ms_mean_ci95_low": 13.16066959580484, "mean_inter_token_latency_ms_min": 8.423491389310232, "mean_inter_token_latency_ms_p25": 12.166923287674537, "mean_inter_token_latency_ms_p50": 13.710881604807572, "mean_inter_token_latency_ms_p75": 14.56516497055845, "mean_inter_token_latency_ms_p90": 15.314808806337037, "mean_inter_token_latency_ms_p95": 16.20452485317018, "mean_inter_token_latency_ms_p99": 18.557768297534793, "mean_inter_token_latency_ms_samples": 100, "mean_inter_token_latency_ms_stddev": 2.0461327741436057, "output_token_sources": { "usage": 100 }, "output_tokens": 50929, "prefix_cache_hit_avg": null, "prompt_token_samples": 100, "prompt_tokens": 365425, "prompt_tokens_max": 3689.0, "prompt_tokens_mean": 3654.25, "prompt_tokens_mean_ci95_high": 3658.3042454864803, "prompt_tokens_mean_ci95_low": 3650.1957545135197, "prompt_tokens_min": 3613.0, "prompt_tokens_p25": 3636.0, "prompt_tokens_p50": 3655.0, "prompt_tokens_p75": 3672.0, "prompt_tokens_p90": 3680.0, "prompt_tokens_p95": 3685.0, "prompt_tokens_p99": 3689.0, "prompt_tokens_samples": 100, "prompt_tokens_stddev": 20.68492595143047, "reasoning_chunks": 0, "request_canary_passed": 100, "request_canary_samples": 100, "throughput_tok_s": 503.3548811864182, "time_to_first_output_max_ms": 3452.846899977885, "time_to_first_output_mean_ci95_high_ms": 1178.8733678174199, "time_to_first_output_mean_ci95_low_ms": 866.4004701862119, "time_to_first_output_mean_ms": 1022.6369190018158, "time_to_first_output_min_ms": 469.35769997071475, "time_to_first_output_p25_ms": 479.2176999617368, "time_to_first_output_p50_ms": 483.2573999883607, "time_to_first_output_p75_ms": 1291.198700084351, "time_to_first_output_p90_ms": 2142.755900043994, "time_to_first_output_p95_ms": 2471.011899993755, "time_to_first_output_p99_ms": 3424.782400019467, "time_to_first_output_samples": 100, "time_to_first_output_stddev_ms": 797.1247388551221, "tpot_max_ms": 22.227747749478908, "tpot_mean_ci95_high_ms": 13.962753643269135, "tpot_mean_ci95_low_ms": 13.16066959580484, "tpot_mean_ms": 13.561711619536988, "tpot_min_ms": 8.423491389310232, "tpot_p25_ms": 12.166923287674537, "tpot_p50_ms": 13.710881604807572, "tpot_p75_ms": 14.56516497055845, "tpot_p90_ms": 15.314808806337037, "tpot_p95_ms": 16.20452485317018, "tpot_p99_ms": 18.557768297534793, "tpot_samples": 100, "tpot_stddev_ms": 2.0461327741436057, "ttft_max_ms": 3452.8471999801695, "ttft_mean_ci95_high_ms": 1178.8737965773878, "ttft_mean_ci95_low_ms": 866.400911428212, "ttft_mean_ms": 1022.6373540027998, "ttft_min_ms": 469.35789997223765, "ttft_p25_ms": 479.2180999647826, "ttft_p50_ms": 483.2577999914065, "ttft_p75_ms": 1291.198900085874, "ttft_p90_ms": 2142.7562000462785, "ttft_p95_ms": 2471.012099995278, "ttft_p99_ms": 3424.78260002099, "ttft_samples": 100, "ttft_stddev_ms": 797.1247070132034 }, "model": "qwen38-27b-kelnei-nvfp4-vllm0271-mtp2", "prompt_cache_mode": "unique", "request_canaries": true, "request_timings": [ { "content_chunks": 173, "decode_tok_s": 85.87239463031749, "e2e_ms": 8347.84189995844, "effective_prefill_tok_s": 1507.2052797386484, "generation_ms": 5950.689999968745, "mean_inter_token_latency_ms": 11.645185909919265, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3613, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00000", "passed": true }, "request_index": 0, "time_to_first_output_ms": 2397.1518999896944, "tpot_ms": 11.645185909919265, "ttft_ms": 2397.15229999274, "visible_content_truncated": false, "visible_generation_ms": 5950.689599965699 }, { "content_chunks": 172, "decode_tok_s": 92.25439834586915, "e2e_ms": 7935.791300027631, "effective_prefill_tok_s": 1508.2861863312023, "generation_ms": 5539.031299995258, "mean_inter_token_latency_ms": 10.839591585117923, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3615, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00001", "passed": true }, "request_index": 1, "time_to_first_output_ms": 2396.760000032373, "tpot_ms": 10.839591585117923, "ttft_ms": 2396.7608000384644, "visible_content_truncated": false, "visible_generation_ms": 5539.030499989167 }, { "content_chunks": 172, "decode_tok_s": 79.64101384293784, "e2e_ms": 7908.379700034857, "effective_prefill_tok_s": 2424.120244506355, "generation_ms": 6416.292000096291, "mean_inter_token_latency_ms": 12.556344422889024, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3617, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00002", "passed": true }, "request_index": 2, "time_to_first_output_ms": 1492.0876999385655, "tpot_ms": 12.556344422889024, "ttft_ms": 1492.0885999454185, "visible_content_truncated": false, "visible_generation_ms": 6416.291100089438 }, { "content_chunks": 172, "decode_tok_s": 101.35048027375713, "e2e_ms": 8347.616899991408, "effective_prefill_tok_s": 1094.1683910455442, "generation_ms": 5041.91000002902, "mean_inter_token_latency_ms": 9.866751467767163, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3617, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00003", "passed": true }, "request_index": 3, "time_to_first_output_ms": 3305.706899962388, "tpot_ms": 9.866751467767163, "ttft_ms": 3305.707299965434, "visible_content_truncated": false, "visible_generation_ms": 5041.909600025974 }, { "content_chunks": 173, "decode_tok_s": 93.61689934876236, "e2e_ms": 8763.958800002001, "effective_prefill_tok_s": 1094.8278204034864, "generation_ms": 5458.416200010106, "mean_inter_token_latency_ms": 10.68183209395324, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3619, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00004", "passed": true }, "request_index": 4, "time_to_first_output_ms": 3305.5425999918953, "tpot_ms": 10.68183209395324, "ttft_ms": 3305.542999994941, "visible_content_truncated": false, "visible_generation_ms": 5458.41580000706 }, { "content_chunks": 172, "decode_tok_s": 95.72553454929647, "e2e_ms": 8762.960999971256, "effective_prefill_tok_s": 1057.001460875127, "generation_ms": 5338.178599951789, "mean_inter_token_latency_ms": 10.446533463702131, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3620, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00005", "passed": true }, "request_index": 5, "time_to_first_output_ms": 3424.782400019467, "tpot_ms": 10.446533463702131, "ttft_ms": 3424.78260002099, "visible_content_truncated": false, "visible_generation_ms": 5338.178399950266 }, { "content_chunks": 172, "decode_tok_s": 95.72367500982321, "e2e_ms": 8762.953200028278, "effective_prefill_tok_s": 1057.6198723256841, "generation_ms": 5338.28230004292, "mean_inter_token_latency_ms": 10.446736399301212, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3622, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00006", "passed": true }, "request_index": 6, "time_to_first_output_ms": 3424.670899985358, "tpot_ms": 10.446736399301212, "ttft_ms": 3424.671099986881, "visible_content_truncated": false, "visible_generation_ms": 5338.282100041397 }, { "content_chunks": 172, "decode_tok_s": 83.42015450455781, "e2e_ms": 9578.46490002703, "effective_prefill_tok_s": 1048.9894585315087, "generation_ms": 6125.618000049144, "mean_inter_token_latency_ms": 11.987510763305565, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3622, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00007", "passed": true }, "request_index": 7, "time_to_first_output_ms": 3452.846899977885, "tpot_ms": 11.987510763305565, "ttft_ms": 3452.8471999801695, "visible_content_truncated": false, "visible_generation_ms": 6125.61770004686 }, { "content_chunks": 174, "decode_tok_s": 73.02972101029377, "e2e_ms": 7852.266900008544, "effective_prefill_tok_s": 4235.6841299942935, "generation_ms": 6997.151199961081, "mean_inter_token_latency_ms": 13.693055185833819, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3622, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00008", "passed": true }, "request_index": 8, "time_to_first_output_ms": 855.1157000474632, "tpot_ms": 13.693055185833819, "ttft_ms": 855.116100050509, "visible_content_truncated": false, "visible_generation_ms": 6997.150799958035 }, { "content_chunks": 172, "decode_tok_s": 83.0198818317061, "e2e_ms": 7799.292800016701, "effective_prefill_tok_s": 2204.1909187039596, "generation_ms": 6155.152100021951, "mean_inter_token_latency_ms": 12.045307436442174, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3624, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00009", "passed": true }, "request_index": 9, "time_to_first_output_ms": 1644.1406999947503, "tpot_ms": 12.045307436442174, "ttft_ms": 1644.1411999985576, "visible_content_truncated": false, "visible_generation_ms": 6155.151600018144 }, { "content_chunks": 173, "decode_tok_s": 89.90889009710614, "e2e_ms": 7826.287299976684, "effective_prefill_tok_s": 1692.213284735584, "generation_ms": 5683.53139993269, "mean_inter_token_latency_ms": 11.122370645660842, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3626, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00010", "passed": true }, "request_index": 10, "time_to_first_output_ms": 2142.755900043994, "tpot_ms": 11.122370645660842, "ttft_ms": 2142.7562000462785, "visible_content_truncated": false, "visible_generation_ms": 5683.531099930406 }, { "content_chunks": 172, "decode_tok_s": 96.9609975606853, "e2e_ms": 7412.28809999302, "effective_prefill_tok_s": 1693.1762895467468, "generation_ms": 5270.1603000750765, "mean_inter_token_latency_ms": 10.313425244765316, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3627, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00011", "passed": true }, "request_index": 11, "time_to_first_output_ms": 2142.127799917944, "tpot_ms": 10.313425244765316, "ttft_ms": 2142.128299921751, "visible_content_truncated": false, "visible_generation_ms": 5270.159800071269 }, { "content_chunks": 172, "decode_tok_s": 102.93765742840458, "e2e_ms": 7409.336299984716, "effective_prefill_tok_s": 1483.3345098423033, "generation_ms": 4964.169700047933, "mean_inter_token_latency_ms": 9.71461780831298, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3627, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00012", "passed": true }, "request_index": 12, "time_to_first_output_ms": 2445.1665999367833, "tpot_ms": 9.71461780831298, "ttft_ms": 2445.167199941352, "visible_content_truncated": false, "visible_generation_ms": 4964.169100043364 }, { "content_chunks": 207, "decode_tok_s": 61.711158399338345, "e2e_ms": 10725.583999999799, "effective_prefill_tok_s": 1483.8010073797132, "generation_ms": 8280.512199969962, "mean_inter_token_latency_ms": 16.20452485317018, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3628, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00013", "passed": true }, "request_index": 13, "time_to_first_output_ms": 2445.0718000298366, "tpot_ms": 16.20452485317018, "ttft_ms": 2445.072100032121, "visible_content_truncated": false, "visible_generation_ms": 8280.511899967678 }, { "content_chunks": 172, "decode_tok_s": 88.70800864937625, "e2e_ms": 8231.483899988234, "effective_prefill_tok_s": 1468.6291069699712, "generation_ms": 5760.471999994479, "mean_inter_token_latency_ms": 11.27293933462716, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3629, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00014", "passed": true }, "request_index": 14, "time_to_first_output_ms": 2471.011899993755, "tpot_ms": 11.27293933462716, "ttft_ms": 2471.012099995278, "visible_content_truncated": false, "visible_generation_ms": 5760.471799992956 }, { "content_chunks": 173, "decode_tok_s": 77.60748191942865, "e2e_ms": 8240.206100046635, "effective_prefill_tok_s": 2192.3077898870433, "generation_ms": 6584.416700061411, "mean_inter_token_latency_ms": 12.88535557741959, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3630, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00015", "passed": true }, "request_index": 15, "time_to_first_output_ms": 1655.789399985224, "tpot_ms": 12.88535557741959, "ttft_ms": 1655.7896999875084, "visible_content_truncated": false, "visible_generation_ms": 6584.416400059126 }, { "content_chunks": 151, "decode_tok_s": 74.79427077890227, "e2e_ms": 7264.434600016102, "effective_prefill_tok_s": 2877.187919411079, "generation_ms": 6003.133599995635, "mean_inter_token_latency_ms": 13.37000801780765, "output_token_source": "usage", "output_tokens": 450, "planned_context_tokens": 4096, "prompt_tokens": 3629, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00016", "passed": true }, "request_index": 16, "time_to_first_output_ms": 1261.3010000204667, "tpot_ms": 13.37000801780765, "ttft_ms": 1261.301500024274, "visible_content_truncated": false, "visible_generation_ms": 6003.133099991828 }, { "content_chunks": 172, "decode_tok_s": 82.96251308875894, "e2e_ms": 8219.621199998073, "effective_prefill_tok_s": 1761.953910740771, "generation_ms": 6159.408399951644, "mean_inter_token_latency_ms": 12.053636790512025, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3630, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00017", "passed": true }, "request_index": 17, "time_to_first_output_ms": 2060.212800046429, "tpot_ms": 12.053636790512025, "ttft_ms": 2060.213000047952, "visible_content_truncated": false, "visible_generation_ms": 6159.408199950121 }, { "content_chunks": 172, "decode_tok_s": 82.95267093904047, "e2e_ms": 8219.337899936363, "effective_prefill_tok_s": 1763.3072515980125, "generation_ms": 6160.139200044796, "mean_inter_token_latency_ms": 12.055066927680619, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3631, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00018", "passed": true }, "request_index": 18, "time_to_first_output_ms": 2059.1986998915672, "tpot_ms": 12.055066927680619, "ttft_ms": 2059.199099894613, "visible_content_truncated": false, "visible_generation_ms": 6160.13880004175 }, { "content_chunks": 172, "decode_tok_s": 88.61955942238079, "e2e_ms": 7832.281099981628, "effective_prefill_tok_s": 1757.9356492139148, "generation_ms": 5766.221400001086, "mean_inter_token_latency_ms": 11.284190606655747, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3632, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00019", "passed": true }, "request_index": 19, "time_to_first_output_ms": 2066.059699980542, "tpot_ms": 11.284190606655747, "ttft_ms": 2066.0599999828264, "visible_content_truncated": false, "visible_generation_ms": 5766.221099998802 }, { "content_chunks": 184, "decode_tok_s": 63.33231646360213, "e2e_ms": 10134.003399987705, "effective_prefill_tok_s": 1758.9363455664538, "generation_ms": 8068.550599971786, "mean_inter_token_latency_ms": 15.789727201510345, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3633, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00020", "passed": true }, "request_index": 20, "time_to_first_output_ms": 2065.452800015919, "tpot_ms": 15.789727201510345, "ttft_ms": 2065.453299903311, "visible_content_truncated": false, "visible_generation_ms": 8068.550100084394 }, { "content_chunks": 172, "decode_tok_s": 82.80474682802334, "e2e_ms": 7832.31920003891, "effective_prefill_tok_s": 2187.607642130451, "generation_ms": 6171.143800020218, "mean_inter_token_latency_ms": 12.076602348376161, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3634, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00021", "passed": true }, "request_index": 21, "time_to_first_output_ms": 1661.175400018692, "tpot_ms": 12.076602348376161, "ttft_ms": 1661.1757000209764, "visible_content_truncated": false, "visible_generation_ms": 6171.143500017934 }, { "content_chunks": 172, "decode_tok_s": 77.81662094488514, "e2e_ms": 7430.869900039397, "effective_prefill_tok_s": 4206.448560701366, "generation_ms": 6566.72050000634, "mean_inter_token_latency_ms": 12.850725048936088, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3635, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00022", "passed": true }, "request_index": 22, "time_to_first_output_ms": 864.1494000330567, "tpot_ms": 12.850725048936088, "ttft_ms": 864.1500000376254, "visible_content_truncated": false, "visible_generation_ms": 6566.719900001772 }, { "content_chunks": 172, "decode_tok_s": 69.4538133089583, "e2e_ms": 7828.844600007869, "effective_prefill_tok_s": 7712.586109478752, "generation_ms": 7357.407400035299, "mean_inter_token_latency_ms": 14.398057534315654, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3636, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00023", "passed": true }, "request_index": 23, "time_to_first_output_ms": 471.43719997256994, "tpot_ms": 14.398057534315654, "ttft_ms": 471.4374999748543, "visible_content_truncated": false, "visible_generation_ms": 7357.407100033015 }, { "content_chunks": 177, "decode_tok_s": 68.21783886307098, "e2e_ms": 7960.067000007257, "effective_prefill_tok_s": 7746.756898260891, "generation_ms": 7490.709300036542, "mean_inter_token_latency_ms": 14.658922309269162, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3636, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00024", "passed": true }, "request_index": 24, "time_to_first_output_ms": 469.35769997071475, "tpot_ms": 14.658922309269162, "ttft_ms": 469.35789997223765, "visible_content_truncated": false, "visible_generation_ms": 7490.709100035019 }, { "content_chunks": 172, "decode_tok_s": 77.91539985593069, "e2e_ms": 7829.413600033149, "effective_prefill_tok_s": 2861.485382150588, "generation_ms": 6558.395399944857, "mean_inter_token_latency_ms": 12.834433267993848, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3637, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00025", "passed": true }, "request_index": 25, "time_to_first_output_ms": 1271.0182000882924, "tpot_ms": 12.834433267993848, "ttft_ms": 1271.0186000913382, "visible_content_truncated": false, "visible_generation_ms": 6558.394999941811 }, { "content_chunks": 207, "decode_tok_s": 53.90497161081138, "e2e_ms": 10750.324600026943, "effective_prefill_tok_s": 2863.8220732543323, "generation_ms": 9479.645100072958, "mean_inter_token_latency_ms": 18.551164579399135, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3639, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00026", "passed": true }, "request_index": 26, "time_to_first_output_ms": 1270.6794999539852, "tpot_ms": 18.551164579399135, "ttft_ms": 1270.6799999577925, "visible_content_truncated": false, "visible_generation_ms": 9479.64460006915 }, { "content_chunks": 173, "decode_tok_s": 77.60905427653579, "e2e_ms": 8249.917500070296, "effective_prefill_tok_s": 2184.7534109206467, "generation_ms": 6584.28329997696, "mean_inter_token_latency_ms": 12.885094520502857, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3639, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00027", "passed": true }, "request_index": 27, "time_to_first_output_ms": 1665.6342000933364, "tpot_ms": 12.885094520502857, "ttft_ms": 1665.6345000956208, "visible_content_truncated": false, "visible_generation_ms": 6584.2829999746755 }, { "content_chunks": 192, "decode_tok_s": 64.75746782199587, "e2e_ms": 9154.595199972391, "effective_prefill_tok_s": 2880.627888397461, "generation_ms": 7890.981800039299, "mean_inter_token_latency_ms": 15.442234442346965, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3640, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00028", "passed": true }, "request_index": 28, "time_to_first_output_ms": 1263.6133999330923, "tpot_ms": 15.442234442346965, "ttft_ms": 1263.6136999353766, "visible_content_truncated": false, "visible_generation_ms": 7890.9815000370145 }, { "content_chunks": 172, "decode_tok_s": 82.90647181242568, "e2e_ms": 7031.688800081611, "effective_prefill_tok_s": 4195.287524183054, "generation_ms": 6163.571900106035, "mean_inter_token_latency_ms": 12.061784540324922, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3642, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00029", "passed": true }, "request_index": 29, "time_to_first_output_ms": 868.1168999755755, "tpot_ms": 12.061784540324922, "ttft_ms": 868.1174999801442, "visible_content_truncated": false, "visible_generation_ms": 6163.5713001014665 }, { "content_chunks": 172, "decode_tok_s": 77.81766851191988, "e2e_ms": 7039.414600003511, "effective_prefill_tok_s": 7703.330813253954, "generation_ms": 6566.632100031711, "mean_inter_token_latency_ms": 12.850552054856578, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3642, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00030", "passed": true }, "request_index": 30, "time_to_first_output_ms": 472.78249997179955, "tpot_ms": 12.850552054856578, "ttft_ms": 472.7829999756068, "visible_content_truncated": false, "visible_generation_ms": 6566.631600027904 }, { "content_chunks": 172, "decode_tok_s": 73.37625983316137, "e2e_ms": 7437.373100081459, "effective_prefill_tok_s": 7697.544602763402, "generation_ms": 6964.1053000232205, "mean_inter_token_latency_ms": 13.628386105720589, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3643, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00031", "passed": true }, "request_index": 31, "time_to_first_output_ms": 473.2678000582382, "tpot_ms": 13.628386105720589, "ttft_ms": 473.268200061284, "visible_content_truncated": false, "visible_generation_ms": 6964.104900020175 }, { "content_chunks": 196, "decode_tok_s": 55.49414396719244, "e2e_ms": 9680.693799979053, "effective_prefill_tok_s": 7711.924488673524, "generation_ms": 9208.178799948655, "mean_inter_token_latency_ms": 18.01991937367643, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3644, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00032", "passed": true }, "request_index": 32, "time_to_first_output_ms": 472.51500003039837, "tpot_ms": 18.01991937367643, "ttft_ms": 472.5155000342056, "visible_content_truncated": false, "visible_generation_ms": 9208.178299944848 }, { "content_chunks": 172, "decode_tok_s": 73.22601103718254, "e2e_ms": 7451.697100070305, "effective_prefill_tok_s": 7701.205886953348, "generation_ms": 6978.394599980675, "mean_inter_token_latency_ms": 13.656349510725391, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3645, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00033", "passed": true }, "request_index": 33, "time_to_first_output_ms": 473.3025000896305, "tpot_ms": 13.656349510725391, "ttft_ms": 473.30280009191483, "visible_content_truncated": false, "visible_generation_ms": 6978.39429997839 }, { "content_chunks": 172, "decode_tok_s": 77.69049008751477, "e2e_ms": 7452.572799986228, "effective_prefill_tok_s": 4165.946824236322, "generation_ms": 6577.381600043736, "mean_inter_token_latency_ms": 12.871588258402616, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3646, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00034", "passed": true }, "request_index": 34, "time_to_first_output_ms": 875.191199942492, "tpot_ms": 12.871588258402616, "ttft_ms": 875.1917999470606, "visible_content_truncated": false, "visible_generation_ms": 6577.381000039168 }, { "content_chunks": 172, "decode_tok_s": 77.6553577113369, "e2e_ms": 7454.869200009853, "effective_prefill_tok_s": 4169.182832119164, "generation_ms": 6580.357300001197, "mean_inter_token_latency_ms": 12.8774115459906, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3646, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00035", "passed": true }, "request_index": 35, "time_to_first_output_ms": 874.5119000086561, "tpot_ms": 12.8774115459906, "ttft_ms": 874.5125000132248, "visible_content_truncated": false, "visible_generation_ms": 6580.356699996628 }, { "content_chunks": 172, "decode_tok_s": 69.29374768217501, "e2e_ms": 7850.366499973461, "effective_prefill_tok_s": 7662.347430923353, "generation_ms": 7374.402699992061, "mean_inter_token_latency_ms": 14.431316438340627, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3647, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00036", "passed": true }, "request_index": 36, "time_to_first_output_ms": 475.96379998140037, "tpot_ms": 14.431316438340627, "ttft_ms": 475.9642999852076, "visible_content_truncated": false, "visible_generation_ms": 7374.402199988253 }, { "content_chunks": 173, "decode_tok_s": 68.97765384504757, "e2e_ms": 7884.27150005009, "effective_prefill_tok_s": 7664.755483770133, "generation_ms": 7408.196300035343, "mean_inter_token_latency_ms": 14.497448728053508, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3649, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00037", "passed": true }, "request_index": 37, "time_to_first_output_ms": 476.0752000147477, "tpot_ms": 14.497448728053508, "ttft_ms": 476.0757000185549, "visible_content_truncated": false, "visible_generation_ms": 7408.195800031535 }, { "content_chunks": 172, "decode_tok_s": 73.1459721916384, "e2e_ms": 7856.05800000485, "effective_prefill_tok_s": 4194.120783001777, "generation_ms": 6986.0306000337005, "mean_inter_token_latency_ms": 13.67129275936145, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3649, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00038", "passed": true }, "request_index": 38, "time_to_first_output_ms": 870.0273999711499, "tpot_ms": 13.67129275936145, "ttft_ms": 870.0276999734342, "visible_content_truncated": false, "visible_generation_ms": 6986.030300031416 }, { "content_chunks": 172, "decode_tok_s": 73.15621673977128, "e2e_ms": 7856.626900029369, "effective_prefill_tok_s": 4187.822820930848, "generation_ms": 6985.052300035022, "mean_inter_token_latency_ms": 13.669378277955033, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3650, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00039", "passed": true }, "request_index": 39, "time_to_first_output_ms": 871.5745999943465, "tpot_ms": 13.669378277955033, "ttft_ms": 871.5750999981537, "visible_content_truncated": false, "visible_generation_ms": 6985.051800031215 }, { "content_chunks": 172, "decode_tok_s": 69.20306508637506, "e2e_ms": 7861.381799913943, "effective_prefill_tok_s": 7649.023980870515, "generation_ms": 7384.065999998711, "mean_inter_token_latency_ms": 14.450227005868319, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3651, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00040", "passed": true }, "request_index": 40, "time_to_first_output_ms": 477.31579991523176, "tpot_ms": 14.450227005868319, "ttft_ms": 477.31619991827756, "visible_content_truncated": false, "visible_generation_ms": 7384.065599995665 }, { "content_chunks": 172, "decode_tok_s": 73.17460715631273, "e2e_ms": 7860.794600099325, "effective_prefill_tok_s": 4161.833795842882, "generation_ms": 6983.296800055541, "mean_inter_token_latency_ms": 13.665942857251547, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3652, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00041", "passed": true }, "request_index": 41, "time_to_first_output_ms": 877.4978000437841, "tpot_ms": 13.665942857251547, "ttft_ms": 877.4982000468299, "visible_content_truncated": false, "visible_generation_ms": 6983.296400052495 }, { "content_chunks": 172, "decode_tok_s": 73.17999665503872, "e2e_ms": 7857.856700080447, "effective_prefill_tok_s": 4174.5031450719325, "generation_ms": 6982.782500097528, "mean_inter_token_latency_ms": 13.664936399408079, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3653, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00042", "passed": true }, "request_index": 42, "time_to_first_output_ms": 875.0741999829188, "tpot_ms": 13.664936399408079, "ttft_ms": 875.0745999859646, "visible_content_truncated": false, "visible_generation_ms": 6982.782100094482 }, { "content_chunks": 172, "decode_tok_s": 69.22668576272476, "e2e_ms": 7857.935000094585, "effective_prefill_tok_s": 7668.1112147909125, "generation_ms": 7381.546500022523, "mean_inter_token_latency_ms": 14.445296477539184, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3653, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00043", "passed": true }, "request_index": 43, "time_to_first_output_ms": 476.388500072062, "tpot_ms": 14.445296477539184, "ttft_ms": 476.3889000751078, "visible_content_truncated": false, "visible_generation_ms": 7381.546100019477 }, { "content_chunks": 172, "decode_tok_s": 69.26612540215545, "e2e_ms": 7858.428900013678, "effective_prefill_tok_s": 7595.325069650224, "generation_ms": 7377.343499916606, "mean_inter_token_latency_ms": 14.437071428408231, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3654, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00044", "passed": true }, "request_index": 44, "time_to_first_output_ms": 481.0854000970721, "tpot_ms": 14.437071428408231, "ttft_ms": 481.0857000993565, "visible_content_truncated": false, "visible_generation_ms": 7377.343199914321 }, { "content_chunks": 172, "decode_tok_s": 69.24314773288494, "e2e_ms": 7855.117000057362, "effective_prefill_tok_s": 7685.261507011951, "generation_ms": 7379.791600047611, "mean_inter_token_latency_ms": 14.441862231012937, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3653, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00045", "passed": true }, "request_index": 45, "time_to_first_output_ms": 475.3254000097513, "tpot_ms": 14.441862231012937, "ttft_ms": 475.32590001355857, "visible_content_truncated": false, "visible_generation_ms": 7379.791100043803 }, { "content_chunks": 172, "decode_tok_s": 73.1818873081247, "e2e_ms": 7856.534600025043, "effective_prefill_tok_s": 4181.100943052084, "generation_ms": 6982.602100004442, "mean_inter_token_latency_ms": 13.664583365957812, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3654, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00046", "passed": true }, "request_index": 46, "time_to_first_output_ms": 873.9325000206009, "tpot_ms": 13.664583365957812, "ttft_ms": 873.9329000236467, "visible_content_truncated": false, "visible_generation_ms": 6982.601700001396 }, { "content_chunks": 172, "decode_tok_s": 73.20053390873382, "e2e_ms": 7855.393799953163, "effective_prefill_tok_s": 4179.194722401193, "generation_ms": 6980.82339996472, "mean_inter_token_latency_ms": 13.66110254396227, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00047", "passed": true }, "request_index": 47, "time_to_first_output_ms": 874.5703999884427, "tpot_ms": 13.66110254396227, "ttft_ms": 874.570699990727, "visible_content_truncated": false, "visible_generation_ms": 6980.823099962436 }, { "content_chunks": 176, "decode_tok_s": 64.66751658329545, "e2e_ms": 8375.649599940516, "effective_prefill_tok_s": 7715.990741238028, "generation_ms": 7901.957999914885, "mean_inter_token_latency_ms": 15.46371428554772, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00048", "passed": true }, "request_index": 48, "time_to_first_output_ms": 473.6916000256315, "tpot_ms": 15.46371428554772, "ttft_ms": 473.69210002943873, "visible_content_truncated": false, "visible_generation_ms": 7901.957499911077 }, { "content_chunks": 172, "decode_tok_s": 77.36223776310982, "e2e_ms": 7480.623799958266, "effective_prefill_tok_s": 4175.548990527875, "generation_ms": 6605.289799976163, "mean_inter_token_latency_ms": 12.926203131068812, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00049", "passed": true }, "request_index": 49, "time_to_first_output_ms": 875.3339999821037, "tpot_ms": 12.926203131068812, "ttft_ms": 875.334499985911, "visible_content_truncated": false, "visible_generation_ms": 6605.289299972355 }, { "content_chunks": 205, "decode_tok_s": 53.88578971173164, "e2e_ms": 10357.136900071055, "effective_prefill_tok_s": 4181.3610139867005, "generation_ms": 9483.01960004028, "mean_inter_token_latency_ms": 18.557768297534793, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00050", "passed": true }, "request_index": 50, "time_to_first_output_ms": 874.1173000307754, "tpot_ms": 18.557768297534793, "ttft_ms": 874.1176000330597, "visible_content_truncated": false, "visible_generation_ms": 9483.019300037995 }, { "content_chunks": 175, "decode_tok_s": 72.12780787894162, "e2e_ms": 7560.76080002822, "effective_prefill_tok_s": 7676.723207755152, "generation_ms": 7084.646199946292, "mean_inter_token_latency_ms": 13.864278277781393, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00051", "passed": true }, "request_index": 51, "time_to_first_output_ms": 476.1146000819281, "tpot_ms": 13.864278277781393, "ttft_ms": 476.11520008649677, "visible_content_truncated": false, "visible_generation_ms": 7084.645599941723 }, { "content_chunks": 179, "decode_tok_s": 71.00587703510517, "e2e_ms": 7673.681700020097, "effective_prefill_tok_s": 7660.9592698815795, "generation_ms": 7196.587400045246, "mean_inter_token_latency_ms": 14.08334129167367, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3655, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00052", "passed": true }, "request_index": 52, "time_to_first_output_ms": 477.0942999748513, "tpot_ms": 14.08334129167367, "ttft_ms": 477.0946999778971, "visible_content_truncated": false, "visible_generation_ms": 7196.5870000422 }, { "content_chunks": 172, "decode_tok_s": 72.93477026665892, "e2e_ms": 7481.942499987781, "effective_prefill_tok_s": 7685.806905725801, "generation_ms": 7006.260500056669, "mean_inter_token_latency_ms": 13.710881604807572, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3656, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00053", "passed": true }, "request_index": 53, "time_to_first_output_ms": 475.68199993111193, "tpot_ms": 13.710881604807572, "ttft_ms": 475.68239993415773, "visible_content_truncated": false, "visible_generation_ms": 7006.260100053623 }, { "content_chunks": 174, "decode_tok_s": 76.69045651746126, "e2e_ms": 7535.472699906677, "effective_prefill_tok_s": 4192.2549551899165, "generation_ms": 6663.149799918756, "mean_inter_token_latency_ms": 13.039432093774474, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3657, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00054", "passed": true }, "request_index": 54, "time_to_first_output_ms": 872.3228999879211, "tpot_ms": 13.039432093774474, "ttft_ms": 872.3234999924898, "visible_content_truncated": false, "visible_generation_ms": 6663.149199914187 }, { "content_chunks": 172, "decode_tok_s": 77.3075922951794, "e2e_ms": 7485.419700038619, "effective_prefill_tok_s": 4178.370501673708, "generation_ms": 6609.958800021559, "mean_inter_token_latency_ms": 12.93534011745902, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3658, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00055", "passed": true }, "request_index": 55, "time_to_first_output_ms": 875.4609000170603, "tpot_ms": 12.93534011745902, "ttft_ms": 875.461500021629, "visible_content_truncated": false, "visible_generation_ms": 6609.95820001699 }, { "content_chunks": 214, "decode_tok_s": 44.98881358879211, "e2e_ms": 11837.91989996098, "effective_prefill_tok_s": 7630.21624056499, "generation_ms": 11358.379099983722, "mean_inter_token_latency_ms": 22.227747749478908, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3659, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00056", "passed": true }, "request_index": 56, "time_to_first_output_ms": 479.54079997725785, "tpot_ms": 22.227747749478908, "ttft_ms": 479.54119998030365, "visible_content_truncated": false, "visible_generation_ms": 11358.378699980676 }, { "content_chunks": 172, "decode_tok_s": 72.74970256942434, "e2e_ms": 7503.431900055148, "effective_prefill_tok_s": 7635.368193342428, "generation_ms": 7024.083700031042, "mean_inter_token_latency_ms": 13.745760665422782, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3660, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00057", "passed": true }, "request_index": 57, "time_to_first_output_ms": 479.3482000241056, "tpot_ms": 13.745760665422782, "ttft_ms": 479.3490000301972, "visible_content_truncated": false, "visible_generation_ms": 7024.0829000249505 }, { "content_chunks": 176, "decode_tok_s": 71.6431353089589, "e2e_ms": 7610.841199988499, "effective_prefill_tok_s": 7654.724862890669, "generation_ms": 7132.57449993398, "mean_inter_token_latency_ms": 13.958071428442231, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3661, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00058", "passed": true }, "request_index": 58, "time_to_first_output_ms": 478.2667000545189, "tpot_ms": 13.958071428442231, "ttft_ms": 478.2672000583261, "visible_content_truncated": false, "visible_generation_ms": 7132.573999930173 }, { "content_chunks": 172, "decode_tok_s": 72.7575075425266, "e2e_ms": 7498.776499996893, "effective_prefill_tok_s": 7700.13353770455, "generation_ms": 7023.330199997872, "mean_inter_token_latency_ms": 13.744286105670982, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3661, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00059", "passed": true }, "request_index": 59, "time_to_first_output_ms": 475.446299999021, "tpot_ms": 13.744286105670982, "ttft_ms": 475.4465000005439, "visible_content_truncated": false, "visible_generation_ms": 7023.329999996349 }, { "content_chunks": 178, "decode_tok_s": 71.05489755655846, "e2e_ms": 7668.942100019194, "effective_prefill_tok_s": 7672.008439492561, "generation_ms": 7191.622499958612, "mean_inter_token_latency_ms": 14.073625244537402, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3662, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00060", "passed": true }, "request_index": 60, "time_to_first_output_ms": 477.31960006058216, "tpot_ms": 14.073625244537402, "ttft_ms": 477.32000006362796, "visible_content_truncated": false, "visible_generation_ms": 7191.622099955566 }, { "content_chunks": 172, "decode_tok_s": 82.19063553338037, "e2e_ms": 7508.45200009644, "effective_prefill_tok_s": 2836.8987668286104, "generation_ms": 6217.253300012089, "mean_inter_token_latency_ms": 12.166836203546163, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3663, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00061", "passed": true }, "request_index": 61, "time_to_first_output_ms": 1291.198700084351, "tpot_ms": 12.166836203546163, "ttft_ms": 1291.198900085874, "visible_content_truncated": false, "visible_generation_ms": 6217.253100010566 }, { "content_chunks": 172, "decode_tok_s": 82.19004725812897, "e2e_ms": 7507.926599937491, "effective_prefill_tok_s": 2839.7010822804364, "generation_ms": 6217.297800001688, "mean_inter_token_latency_ms": 12.166923287674537, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3665, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00062", "passed": true }, "request_index": 62, "time_to_first_output_ms": 1290.6287999358028, "tpot_ms": 12.166923287674537, "ttft_ms": 1290.629499941133, "visible_content_truncated": false, "visible_generation_ms": 6217.297099996358 }, { "content_chunks": 172, "decode_tok_s": 82.21297639874373, "e2e_ms": 7508.464200072922, "effective_prefill_tok_s": 2834.7117844424133, "generation_ms": 6215.5638000695035, "mean_inter_token_latency_ms": 12.1635299414276, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3665, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00063", "passed": true }, "request_index": 63, "time_to_first_output_ms": 1292.9004000034183, "tpot_ms": 12.1635299414276, "ttft_ms": 1292.9006000049412, "visible_content_truncated": false, "visible_generation_ms": 6215.563600067981 }, { "content_chunks": 172, "decode_tok_s": 65.32407664371023, "e2e_ms": 8300.406399997883, "effective_prefill_tok_s": 7671.550495614274, "generation_ms": 7822.536899941042, "mean_inter_token_latency_ms": 15.308291389317107, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3666, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00064", "passed": true }, "request_index": 64, "time_to_first_output_ms": 477.8695000568405, "tpot_ms": 15.308291389317107, "ttft_ms": 477.8700000606477, "visible_content_truncated": false, "visible_generation_ms": 7822.536399937235 }, { "content_chunks": 172, "decode_tok_s": 65.29627712924598, "e2e_ms": 8303.446299978532, "effective_prefill_tok_s": 7678.310814458654, "generation_ms": 7825.867300038226, "mean_inter_token_latency_ms": 15.314808806337037, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3667, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00065", "passed": true }, "request_index": 65, "time_to_first_output_ms": 477.57899994030595, "tpot_ms": 15.314808806337037, "ttft_ms": 477.57939994335175, "visible_content_truncated": false, "visible_generation_ms": 7825.86690003518 }, { "content_chunks": 173, "decode_tok_s": 65.09180173186664, "e2e_ms": 8332.97699992545, "effective_prefill_tok_s": 7597.51806292482, "generation_ms": 7850.451000034809, "mean_inter_token_latency_ms": 15.362917808287298, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3666, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00066", "passed": true }, "request_index": 66, "time_to_first_output_ms": 482.5259998906404, "tpot_ms": 15.362917808287298, "ttft_ms": 482.5264998944476, "visible_content_truncated": false, "visible_generation_ms": 7850.450500031002 }, { "content_chunks": 172, "decode_tok_s": 65.32252343937496, "e2e_ms": 8301.975900074467, "effective_prefill_tok_s": 7651.490966186159, "generation_ms": 7822.722900076769, "mean_inter_token_latency_ms": 15.30865538175493, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3667, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00067", "passed": true }, "request_index": 67, "time_to_first_output_ms": 479.25299999769777, "tpot_ms": 15.30865538175493, "ttft_ms": 479.2532999999821, "visible_content_truncated": false, "visible_generation_ms": 7822.722600074485 }, { "content_chunks": 172, "decode_tok_s": 72.87895534373254, "e2e_ms": 8303.22569992859, "effective_prefill_tok_s": 2839.8898296344573, "generation_ms": 7011.626299936324, "mean_inter_token_latency_ms": 13.721382191656213, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3668, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00068", "passed": true }, "request_index": 68, "time_to_first_output_ms": 1291.5993999922648, "tpot_ms": 13.721382191656213, "ttft_ms": 1291.600099997595, "visible_content_truncated": false, "visible_generation_ms": 7011.625599930994 }, { "content_chunks": 104, "decode_tok_s": 97.25613990825097, "e2e_ms": 4468.927499954589, "effective_prefill_tok_s": 2840.3324420364684, "generation_ms": 3177.1772999782115, "mean_inter_token_latency_ms": 10.282127184395506, "output_token_source": "usage", "output_tokens": 310, "planned_context_tokens": 4096, "prompt_tokens": 3669, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00069", "passed": true }, "request_index": 69, "time_to_first_output_ms": 1291.7501999763772, "tpot_ms": 10.282127184395506, "ttft_ms": 1291.7501999763772, "visible_content_truncated": false, "visible_generation_ms": 3177.1772999782115 }, { "content_chunks": 175, "decode_tok_s": 67.01824473270334, "e2e_ms": 8813.652600045316, "effective_prefill_tok_s": 2836.9015869695863, "generation_ms": 7520.34020004794, "mean_inter_token_latency_ms": 14.92130992073004, "output_token_source": "usage", "output_tokens": 505, "planned_context_tokens": 4096, "prompt_tokens": 3669, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00070", "passed": true }, "request_index": 70, "time_to_first_output_ms": 1293.312399997376, "tpot_ms": 14.92130992073004, "ttft_ms": 1293.3126999996603, "visible_content_truncated": false, "visible_generation_ms": 7520.339900045656 }, { "content_chunks": 172, "decode_tok_s": 68.92239272509337, "e2e_ms": 7892.978599993512, "effective_prefill_tok_s": 7664.315510598567, "generation_ms": 7414.13609997835, "mean_inter_token_latency_ms": 14.509072602697358, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3670, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00071", "passed": true }, "request_index": 71, "time_to_first_output_ms": 478.84250001516193, "tpot_ms": 14.509072602697358, "ttft_ms": 478.84290001820773, "visible_content_truncated": false, "visible_generation_ms": 7414.135699975304 }, { "content_chunks": 180, "decode_tok_s": 66.95917825164297, "e2e_ms": 8111.106999916956, "effective_prefill_tok_s": 7654.429311892852, "generation_ms": 7631.515400018543, "mean_inter_token_latency_ms": 14.934472407081298, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3671, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00072", "passed": true }, "request_index": 72, "time_to_first_output_ms": 479.5915998984128, "tpot_ms": 14.934472407081298, "ttft_ms": 479.59229990374297, "visible_content_truncated": false, "visible_generation_ms": 7631.514700013213 }, { "content_chunks": 172, "decode_tok_s": 68.92580827369758, "e2e_ms": 7896.437800023705, "effective_prefill_tok_s": 7607.696452139964, "generation_ms": 7413.768699974753, "mean_inter_token_latency_ms": 14.508353620302843, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00073", "passed": true }, "request_index": 73, "time_to_first_output_ms": 482.6691000489518, "tpot_ms": 14.508353620302843, "ttft_ms": 482.66960005275905, "visible_content_truncated": false, "visible_generation_ms": 7413.768199970946 }, { "content_chunks": 172, "decode_tok_s": 68.91239252478606, "e2e_ms": 7894.429699983448, "effective_prefill_tok_s": 7662.488260123095, "generation_ms": 7415.212000021711, "mean_inter_token_latency_ms": 14.511178082234268, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00074", "passed": true }, "request_index": 74, "time_to_first_output_ms": 479.2176999617368, "tpot_ms": 14.511178082234268, "ttft_ms": 479.2180999647826, "visible_content_truncated": false, "visible_generation_ms": 7415.211600018665 }, { "content_chunks": 172, "decode_tok_s": 68.9221872818842, "e2e_ms": 7892.801900045015, "effective_prefill_tok_s": 7671.677282885914, "generation_ms": 7414.158200030215, "mean_inter_token_latency_ms": 14.509115851331146, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00075", "passed": true }, "request_index": 75, "time_to_first_output_ms": 478.64370001479983, "tpot_ms": 14.509115851331146, "ttft_ms": 478.6442000186071, "visible_content_truncated": false, "visible_generation_ms": 7414.157700026408 }, { "content_chunks": 172, "decode_tok_s": 68.91358581603853, "e2e_ms": 7894.968900014646, "effective_prefill_tok_s": 7651.828467550846, "generation_ms": 7415.083599975333, "mean_inter_token_latency_ms": 14.510926810127852, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00076", "passed": true }, "request_index": 76, "time_to_first_output_ms": 479.88530003931373, "tpot_ms": 14.510926810127852, "ttft_ms": 479.885800043121, "visible_content_truncated": false, "visible_generation_ms": 7415.083099971525 }, { "content_chunks": 192, "decode_tok_s": 58.29964266260424, "e2e_ms": 9246.078499942087, "effective_prefill_tok_s": 7633.838527115363, "generation_ms": 8765.062299906276, "mean_inter_token_latency_ms": 17.15276379629408, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00077", "passed": true }, "request_index": 77, "time_to_first_output_ms": 481.01620003581047, "tpot_ms": 17.15276379629408, "ttft_ms": 481.01660003885627, "visible_content_truncated": false, "visible_generation_ms": 8765.06189990323 }, { "content_chunks": 173, "decode_tok_s": 72.53839364241402, "e2e_ms": 7523.9699999801815, "effective_prefill_tok_s": 7659.179845786129, "generation_ms": 7044.545299955644, "mean_inter_token_latency_ms": 13.78580293533394, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3672, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00078", "passed": true }, "request_index": 78, "time_to_first_output_ms": 479.42470002453774, "tpot_ms": 13.78580293533394, "ttft_ms": 479.42530002910644, "visible_content_truncated": false, "visible_generation_ms": 7044.544699951075 }, { "content_chunks": 172, "decode_tok_s": 72.80026519220833, "e2e_ms": 7498.59109998215, "effective_prefill_tok_s": 7663.971761146496, "generation_ms": 7019.205200020224, "mean_inter_token_latency_ms": 13.736213698669713, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3674, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00079", "passed": true }, "request_index": 79, "time_to_first_output_ms": 479.38589996192604, "tpot_ms": 13.736213698669713, "ttft_ms": 479.38629996497184, "visible_content_truncated": false, "visible_generation_ms": 7019.204800017178 }, { "content_chunks": 172, "decode_tok_s": 68.67786732299835, "e2e_ms": 7920.57770001702, "effective_prefill_tok_s": 7653.468288233188, "generation_ms": 7440.533900051378, "mean_inter_token_latency_ms": 14.560731702644574, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3674, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00080", "passed": true }, "request_index": 80, "time_to_first_output_ms": 480.0437999656424, "tpot_ms": 14.560731702644574, "ttft_ms": 480.04429996944964, "visible_content_truncated": false, "visible_generation_ms": 7440.5334000475705 }, { "content_chunks": 172, "decode_tok_s": 68.67269509099317, "e2e_ms": 7921.156099997461, "effective_prefill_tok_s": 7655.264383072719, "generation_ms": 7441.094299894758, "mean_inter_token_latency_ms": 14.5618283755279, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3675, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00081", "passed": true }, "request_index": 81, "time_to_first_output_ms": 480.0618001027033, "tpot_ms": 14.5618283755279, "ttft_ms": 480.0623001065105, "visible_content_truncated": false, "visible_generation_ms": 7441.09379989095 }, { "content_chunks": 172, "decode_tok_s": 68.65696351681339, "e2e_ms": 7921.633700025268, "effective_prefill_tok_s": 7676.975587934744, "generation_ms": 7442.799299955368, "mean_inter_token_latency_ms": 14.56516497055845, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3676, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00082", "passed": true }, "request_index": 82, "time_to_first_output_ms": 478.83440006989986, "tpot_ms": 14.56516497055845, "ttft_ms": 478.83480007294565, "visible_content_truncated": false, "visible_generation_ms": 7442.798899952322 }, { "content_chunks": 174, "decode_tok_s": 68.1442009707522, "e2e_ms": 7980.536799994297, "effective_prefill_tok_s": 7632.860449919753, "generation_ms": 7498.8038999727, "mean_inter_token_latency_ms": 14.674763013645205, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3677, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00083", "passed": true }, "request_index": 83, "time_to_first_output_ms": 481.73290002159774, "tpot_ms": 14.674763013645205, "ttft_ms": 481.73330002464354, "visible_content_truncated": false, "visible_generation_ms": 7498.803499969654 }, { "content_chunks": 173, "decode_tok_s": 68.06770517473495, "e2e_ms": 7989.559300011024, "effective_prefill_tok_s": 7623.4413875051605, "generation_ms": 7507.231199997477, "mean_inter_token_latency_ms": 14.691254794515611, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3677, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00084", "passed": true }, "request_index": 84, "time_to_first_output_ms": 482.32810001354665, "tpot_ms": 14.691254794515611, "ttft_ms": 482.32850001659244, "visible_content_truncated": false, "visible_generation_ms": 7507.230799994431 }, { "content_chunks": 175, "decode_tok_s": 67.65865195902046, "e2e_ms": 8032.922300044447, "effective_prefill_tok_s": 7657.656530948075, "generation_ms": 7552.618699963205, "mean_inter_token_latency_ms": 14.78007573378318, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3678, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00085", "passed": true }, "request_index": 85, "time_to_first_output_ms": 480.3036000812426, "tpot_ms": 14.78007573378318, "ttft_ms": 480.30410008504987, "visible_content_truncated": false, "visible_generation_ms": 7552.618199959397 }, { "content_chunks": 172, "decode_tok_s": 72.26571018443778, "e2e_ms": 7553.24889998883, "effective_prefill_tok_s": 7630.847039431296, "generation_ms": 7071.126799914055, "mean_inter_token_latency_ms": 13.837821526250597, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3679, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00086", "passed": true }, "request_index": 86, "time_to_first_output_ms": 482.1221000747755, "tpot_ms": 13.837821526250597, "ttft_ms": 482.122499961406, "visible_content_truncated": false, "visible_generation_ms": 7071.126400027424 }, { "content_chunks": 172, "decode_tok_s": 68.45623500452216, "e2e_ms": 7944.911899976432, "effective_prefill_tok_s": 7662.058258390881, "generation_ms": 7464.623200008646, "mean_inter_token_latency_ms": 14.607873189840793, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3680, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00087", "passed": true }, "request_index": 87, "time_to_first_output_ms": 480.28869996778667, "tpot_ms": 14.607873189840793, "ttft_ms": 480.28929997235537, "visible_content_truncated": false, "visible_generation_ms": 7464.622600004077 }, { "content_chunks": 172, "decode_tok_s": 68.45110159985418, "e2e_ms": 7947.7176000364125, "effective_prefill_tok_s": 7624.3237279451405, "generation_ms": 7465.1830000802875, "mean_inter_token_latency_ms": 14.60896868900252, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3679, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00088", "passed": true }, "request_index": 88, "time_to_first_output_ms": 482.534599956125, "tpot_ms": 14.60896868900252, "ttft_ms": 482.5349999591708, "visible_content_truncated": false, "visible_generation_ms": 7465.182600077242 }, { "content_chunks": 172, "decode_tok_s": 68.35978834747843, "e2e_ms": 7955.467200023122, "effective_prefill_tok_s": 7661.68019115898, "generation_ms": 7475.154800107703, "mean_inter_token_latency_ms": 14.628482974770455, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3680, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00089", "passed": true }, "request_index": 89, "time_to_first_output_ms": 480.3123999154195, "tpot_ms": 14.628482974770455, "ttft_ms": 480.3127999184653, "visible_content_truncated": false, "visible_generation_ms": 7475.154400104657 }, { "content_chunks": 172, "decode_tok_s": 68.38629423526528, "e2e_ms": 7955.5149000370875, "effective_prefill_tok_s": 7617.058735341988, "generation_ms": 7472.257500048727, "mean_inter_token_latency_ms": 14.622813111641344, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3681, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00090", "passed": true }, "request_index": 90, "time_to_first_output_ms": 483.2573999883607, "tpot_ms": 14.622813111641344, "ttft_ms": 483.2577999914065, "visible_content_truncated": false, "visible_generation_ms": 7472.257100045681 }, { "content_chunks": 172, "decode_tok_s": 68.43575274947179, "e2e_ms": 7955.527800018899, "effective_prefill_tok_s": 7534.729434322371, "generation_ms": 7466.857300023548, "mean_inter_token_latency_ms": 14.612245205525534, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3682, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00091", "passed": true }, "request_index": 91, "time_to_first_output_ms": 488.67049999535084, "tpot_ms": 14.612245205525534, "ttft_ms": 488.6709999991581, "visible_content_truncated": false, "visible_generation_ms": 7466.856800019741 }, { "content_chunks": 172, "decode_tok_s": 68.63278324792834, "e2e_ms": 7918.560000020079, "effective_prefill_tok_s": 7784.190041522781, "generation_ms": 7445.4215000150725, "mean_inter_token_latency_ms": 14.570296477524604, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3683, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00092", "passed": true }, "request_index": 92, "time_to_first_output_ms": 473.1385000050068, "tpot_ms": 14.570296477524604, "ttft_ms": 473.13920001033694, "visible_content_truncated": false, "visible_generation_ms": 7445.420800009742 }, { "content_chunks": 172, "decode_tok_s": 76.95314923536567, "e2e_ms": 7519.063899992034, "effective_prefill_tok_s": 4192.748525680224, "generation_ms": 6640.404000063427, "mean_inter_token_latency_ms": 12.994919765290463, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3684, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00093", "passed": true }, "request_index": 93, "time_to_first_output_ms": 878.659899928607, "tpot_ms": 12.994919765290463, "ttft_ms": 878.6602999316528, "visible_content_truncated": false, "visible_generation_ms": 6640.403600060381 }, { "content_chunks": 172, "decode_tok_s": 76.98271506079995, "e2e_ms": 7519.330300041474, "effective_prefill_tok_s": 4180.485335285764, "generation_ms": 6637.853699969128, "mean_inter_token_latency_ms": 12.989928962757588, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3685, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00094", "passed": true }, "request_index": 94, "time_to_first_output_ms": 881.4766000723466, "tpot_ms": 12.989928962757588, "ttft_ms": 881.4772000769153, "visible_content_truncated": false, "visible_generation_ms": 6637.853099964559 }, { "content_chunks": 172, "decode_tok_s": 82.15547410583004, "e2e_ms": 6702.697400003672, "effective_prefill_tok_s": 7634.896989173416, "generation_ms": 6219.914200017229, "mean_inter_token_latency_ms": 12.172043444260721, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3686, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00095", "passed": true }, "request_index": 95, "time_to_first_output_ms": 482.7831999864429, "tpot_ms": 12.172043444260721, "ttft_ms": 482.7835999894887, "visible_content_truncated": false, "visible_generation_ms": 6219.913800014183 }, { "content_chunks": 172, "decode_tok_s": 90.69399721042299, "e2e_ms": 6115.344200050458, "effective_prefill_tok_s": 7662.99136884525, "generation_ms": 5634.331000037491, "mean_inter_token_latency_ms": 11.026088062695678, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3686, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00096", "passed": true }, "request_index": 96, "time_to_first_output_ms": 481.013200012967, "tpot_ms": 11.026088062695678, "ttft_ms": 481.01370001677424, "visible_content_truncated": false, "visible_generation_ms": 5634.330500033684 }, { "content_chunks": 172, "decode_tok_s": 99.00612766534235, "e2e_ms": 5641.692199977115, "effective_prefill_tok_s": 7674.926181157242, "generation_ms": 5161.2967000110075, "mean_inter_token_latency_ms": 10.10038493152839, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3687, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00097", "passed": true }, "request_index": 97, "time_to_first_output_ms": 480.3954999661073, "tpot_ms": 10.10038493152839, "ttft_ms": 480.3958999691531, "visible_content_truncated": false, "visible_generation_ms": 5161.296300007962 }, { "content_chunks": 173, "decode_tok_s": 107.54471645148837, "e2e_ms": 5235.124300001189, "effective_prefill_tok_s": 7628.009537996882, "generation_ms": 4751.511899987236, "mean_inter_token_latency_ms": 9.298457729916313, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3689, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00098", "passed": true }, "request_index": 98, "time_to_first_output_ms": 483.61240001395345, "tpot_ms": 9.298457729916313, "ttft_ms": 483.61280001699924, "visible_content_truncated": false, "visible_generation_ms": 4751.51149998419 }, { "content_chunks": 173, "decode_tok_s": 118.71561966206109, "e2e_ms": 4784.955199924298, "effective_prefill_tok_s": 7676.602967096667, "generation_ms": 4304.404099937528, "mean_inter_token_latency_ms": 8.423491389310232, "output_token_source": "usage", "output_tokens": 512, "planned_context_tokens": 4096, "prompt_tokens": 3689, "reasoning_chunks": 0, "request_canary": { "foreign_marker_count": 0, "marker": "ANVIL_REQ_0_00099", "passed": true }, "request_index": 99, "time_to_first_output_ms": 480.55109998676926, "tpot_ms": 8.423491389310232, "ttft_ms": 480.55149998981506, "visible_content_truncated": false, "visible_generation_ms": 4304.403699934483 } ], "requests": 100, "response_words": 256, "run_id": "benchmark-20260905T050535Z", "schema": "anvil-serving.benchmark/v1", "serve_flags": { "no_thinking": false, "reasoning_effort": null, "shared_prefix_burst": false, "thinking_mode": "disabled" }, "source_recipe": null, "started_at": "2026-09-05T05:05:35Z", "tier": null, "timing_methodology": { "clock": "client time.perf_counter", "decode": "usage completion tokens after the first token divided by client-observed generation time; completion tokens may include reasoning tokens", "distributions": "request-level mean, sample standard deviation, normal-approximation 95% confidence interval for the mean, min, p25, p50, p75, p90, p95, p99, and max using the nearest-rank percentile method", "effective_prefill": "usage.prompt_tokens divided by client-observed time to first output; includes queueing, scheduling, prefill, and first-token work", "generation": "client-observed E2E minus time to first output", "mean_inter_token_latency": "client-observed generation time divided by completion tokens after the first token; not a raw per-token timestamp trace", "request_canaries": "when enabled, each response must contain its own synthetic marker and no marker assigned to another request; failures are excluded from performance metrics", "time_to_first_output": "request start through the first non-empty streamed reasoning or content delta", "tpot": "alias of mean_inter_token_latency for capacity-v3: client-observed generation time divided by completion tokens after the first token", "ttft": "request start through the first non-empty streamed content delta" }, "wall_clock_ms": 101179.11220004316 }