{ "base_url": "http://127.0.0.1:39280/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": "2026-09-05T05:19:27Z", "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": 510.35, "completion_tokens_mean_ci95_high": 512.8591549464819, "completion_tokens_mean_ci95_low": 507.84084505351814, "completion_tokens_min": 387.0, "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": 12.801810951438224, "content_chunks": 51032, "content_chunks_s": null, "decode_tok_s_max": 54.315552902738844, "decode_tok_s_mean": 45.641079803364775, "decode_tok_s_mean_ci95_high": 46.1529497813451, "decode_tok_s_mean_ci95_low": 45.12920982538445, "decode_tok_s_min": 41.214421066327986, "decode_tok_s_p25": 43.40927657435038, "decode_tok_s_p50": 46.21114714124146, "decode_tok_s_p75": 47.183004286755846, "decode_tok_s_p90": 48.43357524233903, "decode_tok_s_p95": 48.98024378559647, "decode_tok_s_p99": 54.17163288593884, "decode_tok_s_samples": 100, "decode_tok_s_stddev": 2.611581520307796, "e2e_max_ms": 14372.368199983612, "e2e_mean_ci95_high_ms": 12652.01027233229, "e2e_mean_ci95_low_ms": 12431.754759663427, "e2e_mean_ms": 12541.882515997859, "e2e_min_ms": 9870.348499971442, "e2e_p25_ms": 12538.081500097178, "e2e_p50_ms": 12617.350200074725, "e2e_p75_ms": 12645.580200012773, "e2e_p90_ms": 13014.036499895155, "e2e_p95_ms": 13035.544399986975, "e2e_p99_ms": 13608.274799888022, "e2e_samples": 100, "e2e_stddev_ms": 561.8763078287355, "effective_prefill_tok_s_max": 8599.656772275153, "effective_prefill_tok_s_mean": 3762.4336073274585, "effective_prefill_tok_s_mean_ci95_high": 4247.299458079713, "effective_prefill_tok_s_mean_ci95_low": 3277.5677565752035, "effective_prefill_tok_s_min": 1148.7986347799865, "effective_prefill_tok_s_p25": 2054.7515552527616, "effective_prefill_tok_s_p50": 2324.1874206070006, "effective_prefill_tok_s_p75": 4516.045558704667, "effective_prefill_tok_s_p90": 8442.020436982579, "effective_prefill_tok_s_p95": 8450.376638007341, "effective_prefill_tok_s_p99": 8526.50123145153, "effective_prefill_tok_s_samples": 100, "effective_prefill_tok_s_stddev": 2473.805360980892, "generation_max_ms": 12398.572799982503, "generation_mean_ci95_high_ms": 11344.432516935969, "generation_mean_ci95_low_ms": 11055.598663059307, "generation_mean_ms": 11200.015589997638, "generation_min_ms": 7106.62009997759, "generation_p25_ms": 10781.09000006225, "generation_p50_ms": 11057.30870005209, "generation_p75_ms": 11743.648900068365, "generation_p90_ms": 12197.918700054288, "generation_p95_ms": 12207.628499949351, "generation_p99_ms": 12217.487399931997, "generation_samples": 100, "generation_stddev_ms": 736.8210558078105, "mean_inter_token_latency_ms_max": 24.263351859065565, "mean_inter_token_latency_ms_mean": 21.979798550352697, "mean_inter_token_latency_ms_mean_ci95_high": 22.22186708840106, "mean_inter_token_latency_ms_mean_ci95_low": 21.737730012304336, "mean_inter_token_latency_ms_min": 18.41093290149635, "mean_inter_token_latency_ms_p25": 21.169239725988657, "mean_inter_token_latency_ms_p50": 21.63887788644589, "mean_inter_token_latency_ms_p75": 22.98170039152322, "mean_inter_token_latency_ms_p90": 23.870682387581777, "mean_inter_token_latency_ms_p95": 23.88968395293415, "mean_inter_token_latency_ms_p99": 23.908977299279837, "mean_inter_token_latency_ms_samples": 100, "mean_inter_token_latency_ms_stddev": 1.2350435614712358, "output_token_sources": { "usage": 100 }, "output_tokens": 51035, "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": 315.1821862781392, "time_to_first_output_max_ms": 3152.858900022693, "time_to_first_output_mean_ci95_high_ms": 1468.3902333750214, "time_to_first_output_mean_ci95_low_ms": 1215.34361862542, "time_to_first_output_mean_ms": 1341.8669260002207, "time_to_first_output_min_ms": 421.1796000599861, "time_to_first_output_p25_ms": 799.6942000463605, "time_to_first_output_p50_ms": 1554.973699967377, "time_to_first_output_p75_ms": 1764.9796000914648, "time_to_first_output_p90_ms": 1968.706400017254, "time_to_first_output_p95_ms": 1977.894900017418, "time_to_first_output_p99_ms": 3102.177699911408, "time_to_first_output_samples": 100, "time_to_first_output_stddev_ms": 645.5270784428603, "tpot_max_ms": 24.263351859065565, "tpot_mean_ci95_high_ms": 22.22186708840106, "tpot_mean_ci95_low_ms": 21.737730012304336, "tpot_mean_ms": 21.979798550352697, "tpot_min_ms": 18.41093290149635, "tpot_p25_ms": 21.169239725988657, "tpot_p50_ms": 21.63887788644589, "tpot_p75_ms": 22.98170039152322, "tpot_p90_ms": 23.870682387581777, "tpot_p95_ms": 23.88968395293415, "tpot_p99_ms": 23.908977299279837, "tpot_samples": 100, "tpot_stddev_ms": 1.2350435614712358, "ttft_max_ms": 3152.8594000265, "ttft_mean_ci95_high_ms": 1468.3906984182013, "ttft_mean_ci95_low_ms": 1215.3441115872065, "ttft_mean_ms": 1341.867405002704, "ttft_min_ms": 421.18010006379336, "ttft_p25_ms": 799.6949000516906, "ttft_p50_ms": 1554.9738999688998, "ttft_p75_ms": 1764.9798000929877, "ttft_p90_ms": 1968.7067000195384, "ttft_p95_ms": 1977.8951000189409, "ttft_p99_ms": 3102.178099914454, "ttft_samples": 100, "ttft_stddev_ms": 645.527007221926 }, "model": "qwen38-27b-kelnei-nvfp4-vllm0271-nospec", "prompt_cache_mode": "unique", "request_canaries": true, "request_timings": [ { "content_chunks": 512, "decode_tok_s": 42.868898067670614, "e2e_ms": 12815.384999965318, "effective_prefill_tok_s": 4035.423264768647, "generation_ms": 11920.063799945638, "mean_inter_token_latency_ms": 23.326935029247824, "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": 895.32120001968, "tpot_ms": 23.326935029247824, "ttft_ms": 895.3220000257716, "visible_content_truncated": false, "visible_generation_ms": 11920.062999939546 }, { "content_chunks": 512, "decode_tok_s": 48.638423492310714, "e2e_ms": 13608.274799888022, "effective_prefill_tok_s": 1165.3104205162836, "generation_ms": 10506.097099976614, "mean_inter_token_latency_ms": 20.559876907977717, "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": 3102.177699911408, "tpot_ms": 20.559876907977717, "ttft_ms": 3102.178099914454, "visible_content_truncated": false, "visible_generation_ms": 10506.096699973568 }, { "content_chunks": 512, "decode_tok_s": 46.17161336325688, "e2e_ms": 12833.166499971412, "effective_prefill_tok_s": 2048.4094019079716, "generation_ms": 11067.406199988909, "mean_inter_token_latency_ms": 21.658329158491014, "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": 1765.760299982503, "tpot_ms": 21.658329158491014, "ttft_ms": 1765.7608999870718, "visible_content_truncated": false, "visible_generation_ms": 11067.40559998434 }, { "content_chunks": 512, "decode_tok_s": 48.980944735415555, "e2e_ms": 13044.956999947317, "effective_prefill_tok_s": 1384.5885002481805, "generation_ms": 10432.628499926068, "mean_inter_token_latency_ms": 20.416102739581348, "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": 2612.328500021249, "tpot_ms": 20.416102739581348, "ttft_ms": 2612.328900024295, "visible_content_truncated": false, "visible_generation_ms": 10432.628099923022 }, { "content_chunks": 512, "decode_tok_s": 46.17268346874482, "e2e_ms": 12832.129300106317, "effective_prefill_tok_s": 2050.4486283084834, "generation_ms": 11067.149700014852, "mean_inter_token_latency_ms": 21.65782720159462, "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": 1764.9796000914648, "tpot_ms": 21.65782720159462, "ttft_ms": 1764.9798000929877, "visible_content_truncated": false, "visible_generation_ms": 11067.149500013329 }, { "content_chunks": 512, "decode_tok_s": 48.98024378559647, "e2e_ms": 13044.613399892114, "effective_prefill_tok_s": 1385.9984142988208, "generation_ms": 10432.777799898759, "mean_inter_token_latency_ms": 20.416394911739253, "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": 2611.8355999933556, "tpot_ms": 20.416394911739253, "ttft_ms": 2611.8357999948785, "visible_content_truncated": false, "visible_generation_ms": 10432.777599897236 }, { "content_chunks": 386, "decode_tok_s": 54.315552902738844, "e2e_ms": 10207.277499954216, "effective_prefill_tok_s": 1168.1393758714858, "generation_ms": 7106.62009997759, "mean_inter_token_latency_ms": 18.41093290149635, "output_token_source": "usage", "output_tokens": 387, "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": 3100.657399976626, "tpot_ms": 18.41093290149635, "ttft_ms": 3100.6576999789104, "visible_content_truncated": false, "visible_generation_ms": 7106.619799975306 }, { "content_chunks": 512, "decode_tok_s": 45.545663926833235, "e2e_ms": 14372.368199983612, "effective_prefill_tok_s": 1148.7986347799865, "generation_ms": 11219.509299960919, "mean_inter_token_latency_ms": 21.95598688837753, "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": 3152.858900022693, "tpot_ms": 21.95598688837753, "ttft_ms": 3152.8594000265, "visible_content_truncated": false, "visible_generation_ms": 11219.508799957111 }, { "content_chunks": 512, "decode_tok_s": 42.61521235325161, "e2e_ms": 12412.202800041996, "effective_prefill_tok_s": 8599.656772275153, "generation_ms": 11991.02319998201, "mean_inter_token_latency_ms": 23.465798825796497, "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": 421.1796000599861, "tpot_ms": 23.465798825796497, "ttft_ms": 421.18010006379336, "visible_content_truncated": false, "visible_generation_ms": 11991.022699978203 }, { "content_chunks": 512, "decode_tok_s": 43.9290764291125, "e2e_ms": 12425.172899966128, "effective_prefill_tok_s": 4571.21925504847, "generation_ms": 11632.386599900201, "mean_inter_token_latency_ms": 22.76396594892407, "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": 792.7863000659272, "tpot_ms": 22.76396594892407, "ttft_ms": 792.7866000682116, "visible_content_truncated": false, "visible_generation_ms": 11632.386299897917 }, { "content_chunks": 512, "decode_tok_s": 46.94795135104262, "e2e_ms": 12425.177299999632, "effective_prefill_tok_s": 2353.348456023741, "generation_ms": 10884.394000051543, "mean_inter_token_latency_ms": 21.300183953134134, "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": 1540.7832999480888, "tpot_ms": 21.300183953134134, "ttft_ms": 1540.7836999511346, "visible_content_truncated": false, "visible_generation_ms": 10884.393600048497 }, { "content_chunks": 512, "decode_tok_s": 46.946995968794596, "e2e_ms": 12425.24350003805, "effective_prefill_tok_s": 2354.2347665311854, "generation_ms": 10884.615499991924, "mean_inter_token_latency_ms": 21.30061741681394, "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": 1540.6280000461265, "tpot_ms": 21.30061741681394, "ttft_ms": 1540.628300048411, "visible_content_truncated": false, "visible_generation_ms": 10884.61519998964 }, { "content_chunks": 512, "decode_tok_s": 48.62259977272163, "e2e_ms": 12427.71509999875, "effective_prefill_tok_s": 1890.8362423199446, "generation_ms": 10509.516200050712, "mean_inter_token_latency_ms": 20.566567906165776, "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": 1918.1988999480382, "tpot_ms": 20.566567906165776, "ttft_ms": 1918.1994999526069, "visible_content_truncated": false, "visible_generation_ms": 10509.515600046143 }, { "content_chunks": 512, "decode_tok_s": 48.62635634329601, "e2e_ms": 12426.83130002115, "effective_prefill_tok_s": 1891.4284612191211, "generation_ms": 10508.704300038517, "mean_inter_token_latency_ms": 20.564979060740736, "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": 1918.1269999826327, "tpot_ms": 20.564979060740736, "ttft_ms": 1918.1271999841556, "visible_content_truncated": false, "visible_generation_ms": 10508.704100036994 }, { "content_chunks": 512, "decode_tok_s": 47.8041326047399, "e2e_ms": 12432.116699987091, "effective_prefill_tok_s": 2082.443531984203, "generation_ms": 10689.452399965376, "mean_inter_token_latency_ms": 20.918693542006604, "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": 1742.6643000217155, "tpot_ms": 20.918693542006604, "ttft_ms": 1742.6648000255227, "visible_content_truncated": false, "visible_generation_ms": 10689.451899961568 }, { "content_chunks": 512, "decode_tok_s": 45.33692633631594, "e2e_ms": 12439.295200048946, "effective_prefill_tok_s": 3107.53163795945, "generation_ms": 11271.165500045754, "mean_inter_token_latency_ms": 22.05707534255529, "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": 1168.1297000031918, "tpot_ms": 22.05707534255529, "ttft_ms": 1168.1301000062376, "visible_content_truncated": false, "visible_generation_ms": 11271.165100042708 }, { "content_chunks": 512, "decode_tok_s": 42.413670170025775, "e2e_ms": 12473.616599920206, "effective_prefill_tok_s": 8526.50123145153, "generation_ms": 12048.002399969846, "mean_inter_token_latency_ms": 23.577304109530033, "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_00016", "passed": true }, "request_index": 16, "time_to_first_output_ms": 425.61419995035976, "tpot_ms": 23.577304109530033, "ttft_ms": 425.6144999526441, "visible_content_truncated": false, "visible_generation_ms": 12048.002099967562 }, { "content_chunks": 512, "decode_tok_s": 43.7316843621128, "e2e_ms": 12484.585999976844, "effective_prefill_tok_s": 4539.235122362472, "generation_ms": 11684.891799930483, "mean_inter_token_latency_ms": 22.866715851135975, "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": 799.6942000463605, "tpot_ms": 22.866715851135975, "ttft_ms": 799.6949000516906, "visible_content_truncated": false, "visible_generation_ms": 11684.891099925153 }, { "content_chunks": 512, "decode_tok_s": 46.75198412738607, "e2e_ms": 12484.991100034676, "effective_prefill_tok_s": 2335.087725326916, "generation_ms": 10930.0174000673, "mean_inter_token_latency_ms": 21.389466536335224, "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": 1554.973699967377, "tpot_ms": 21.389466536335224, "ttft_ms": 1554.9738999688998, "visible_content_truncated": false, "visible_generation_ms": 10930.017200065777 }, { "content_chunks": 512, "decode_tok_s": 46.752653549153585, "e2e_ms": 12484.55290007405, "effective_prefill_tok_s": 2336.1540420353044, "generation_ms": 10929.860900039785, "mean_inter_token_latency_ms": 21.389160274050457, "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": 1554.6920000342652, "tpot_ms": 21.389160274050457, "ttft_ms": 1554.6925000380725, "visible_content_truncated": false, "visible_generation_ms": 10929.860400035977 }, { "content_chunks": 512, "decode_tok_s": 48.43443645972095, "e2e_ms": 12485.462299897335, "effective_prefill_tok_s": 1877.4052802687397, "generation_ms": 10550.34469999373, "mean_inter_token_latency_ms": 20.6464671232754, "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": 1935.1175999036059, "tpot_ms": 20.6464671232754, "ttft_ms": 1935.118499910459, "visible_content_truncated": false, "visible_generation_ms": 10550.343799986877 }, { "content_chunks": 512, "decode_tok_s": 48.43357524233903, "e2e_ms": 12485.40659993887, "effective_prefill_tok_s": 1878.158183279497, "generation_ms": 10550.532300025225, "mean_inter_token_latency_ms": 20.646834246624707, "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": 1934.8742999136448, "tpot_ms": 20.646834246624707, "ttft_ms": 1934.8745999159291, "visible_content_truncated": false, "visible_generation_ms": 10550.53200002294 }, { "content_chunks": 512, "decode_tok_s": 47.61059741469938, "e2e_ms": 12490.270799957216, "effective_prefill_tok_s": 2068.4362769957306, "generation_ms": 10732.90459997952, "mean_inter_token_latency_ms": 21.00372720152548, "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": 1757.366199977696, "tpot_ms": 21.00372720152548, "ttft_ms": 1757.3664999799803, "visible_content_truncated": false, "visible_generation_ms": 10732.904299977235 }, { "content_chunks": 512, "decode_tok_s": 45.15042708612222, "e2e_ms": 12495.669199968688, "effective_prefill_tok_s": 3086.726836855577, "generation_ms": 11317.72239995189, "mean_inter_token_latency_ms": 22.148184735717983, "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": 1177.946800016798, "tpot_ms": 22.148184735717983, "ttft_ms": 1177.9473000206053, "visible_content_truncated": false, "visible_generation_ms": 11317.721899948083 }, { "content_chunks": 512, "decode_tok_s": 42.20052289504772, "e2e_ms": 12538.081500097178, "effective_prefill_tok_s": 8471.046074349177, "generation_ms": 12108.8546999963, "mean_inter_token_latency_ms": 23.696388845393933, "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": 429.2268001008779, "tpot_ms": 23.696388845393933, "ttft_ms": 429.22710010316223, "visible_content_truncated": false, "visible_generation_ms": 12108.854399994016 }, { "content_chunks": 512, "decode_tok_s": 43.5128812474141, "e2e_ms": 12548.999499995261, "effective_prefill_tok_s": 4516.045558704667, "generation_ms": 11743.648900068365, "mean_inter_token_latency_ms": 22.98170039152322, "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": 805.3505999268964, "tpot_ms": 22.98170039152322, "ttft_ms": 805.3512999322265, "visible_content_truncated": false, "visible_generation_ms": 11743.648200063035 }, { "content_chunks": 512, "decode_tok_s": 46.52337086077015, "e2e_ms": 12549.534699996002, "effective_prefill_tok_s": 2324.0400257959664, "generation_ms": 10983.72689995449, "mean_inter_token_latency_ms": 21.494573189734815, "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": 1565.8078000415117, "tpot_ms": 21.494573189734815, "ttft_ms": 1565.8084000460804, "visible_content_truncated": false, "visible_generation_ms": 10983.726299949922 }, { "content_chunks": 512, "decode_tok_s": 46.52472122891436, "e2e_ms": 12549.116600072011, "effective_prefill_tok_s": 2324.1874206070006, "generation_ms": 10983.408100088127, "mean_inter_token_latency_ms": 21.493949315240954, "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": 1565.7084999838844, "tpot_ms": 21.493949315240954, "ttft_ms": 1565.7087999861687, "visible_content_truncated": false, "visible_generation_ms": 10983.407800085843 }, { "content_chunks": 512, "decode_tok_s": 48.20360266015685, "e2e_ms": 12550.18919997383, "effective_prefill_tok_s": 1867.31610960682, "generation_ms": 10600.867399945855, "mean_inter_token_latency_ms": 20.745337377584843, "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": 1949.321800027974, "tpot_ms": 20.745337377584843, "ttft_ms": 1949.3220000294968, "visible_content_truncated": false, "visible_generation_ms": 10600.867199944332 }, { "content_chunks": 512, "decode_tok_s": 48.20081497074188, "e2e_ms": 12550.486500025727, "effective_prefill_tok_s": 1868.6448374660717, "generation_ms": 10601.480500074103, "mean_inter_token_latency_ms": 20.746537182141104, "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": 1949.0059999516234, "tpot_ms": 20.746537182141104, "ttft_ms": 1949.0063999546692, "visible_content_truncated": false, "visible_generation_ms": 10601.480100071058 }, { "content_chunks": 512, "decode_tok_s": 47.397804859902806, "e2e_ms": 12551.027200068347, "effective_prefill_tok_s": 2057.700126302477, "generation_ms": 10781.09000006225, "mean_inter_token_latency_ms": 21.09802348348777, "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": 1769.9372000060976, "tpot_ms": 21.09802348348777, "ttft_ms": 1769.9378000106663, "visible_content_truncated": false, "visible_generation_ms": 10781.08940005768 }, { "content_chunks": 512, "decode_tok_s": 44.953739918242135, "e2e_ms": 12552.620599977672, "effective_prefill_tok_s": 3073.2773767806916, "generation_ms": 11367.24110005889, "mean_inter_token_latency_ms": 22.245090215379435, "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": 1185.3794999187812, "tpot_ms": 22.245090215379435, "ttft_ms": 1185.379899921827, "visible_content_truncated": false, "visible_generation_ms": 11367.240700055845 }, { "content_chunks": 512, "decode_tok_s": 42.08480688163572, "e2e_ms": 12573.541000019759, "effective_prefill_tok_s": 8447.07561622416, "generation_ms": 12142.149099963717, "mean_inter_token_latency_ms": 23.761544226934866, "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": 431.3919000560418, "tpot_ms": 23.761544226934866, "ttft_ms": 431.3946000766009, "visible_content_truncated": false, "visible_generation_ms": 12142.146399943158 }, { "content_chunks": 512, "decode_tok_s": 43.40927657435038, "e2e_ms": 12578.969300026074, "effective_prefill_tok_s": 4515.09546895834, "generation_ms": 11771.677399985492, "mean_inter_token_latency_ms": 23.036550684903116, "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": 807.2919000405818, "tpot_ms": 23.036550684903116, "ttft_ms": 807.2925000451505, "visible_content_truncated": false, "visible_generation_ms": 11771.676799980924 }, { "content_chunks": 512, "decode_tok_s": 46.413397867214414, "e2e_ms": 12578.87190009933, "effective_prefill_tok_s": 2323.5955391308908, "generation_ms": 11009.75200009998, "mean_inter_token_latency_ms": 21.545502935616398, "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": 1569.1198999993503, "tpot_ms": 21.545502935616398, "ttft_ms": 1569.1202000016347, "visible_content_truncated": false, "visible_generation_ms": 11009.751700097695 }, { "content_chunks": 512, "decode_tok_s": 46.41558294835977, "e2e_ms": 12578.397799981758, "effective_prefill_tok_s": 2323.5300884279627, "generation_ms": 11009.233699995093, "mean_inter_token_latency_ms": 21.544488649696856, "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": 1569.164099986665, "tpot_ms": 21.544488649696856, "ttft_ms": 1569.1645999904722, "visible_content_truncated": false, "visible_generation_ms": 11009.233199991286 }, { "content_chunks": 512, "decode_tok_s": 48.09894927129347, "e2e_ms": 12579.662199947052, "effective_prefill_tok_s": 1864.7773120669258, "generation_ms": 10623.932700022124, "mean_inter_token_latency_ms": 20.790474951119617, "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": 1955.729499924928, "tpot_ms": 20.790474951119617, "ttft_ms": 1955.7297999272123, "visible_content_truncated": false, "visible_generation_ms": 10623.93240001984 }, { "content_chunks": 512, "decode_tok_s": 48.099538295430044, "e2e_ms": 12578.82949989289, "effective_prefill_tok_s": 1866.4704818795515, "generation_ms": 10623.8025999628, "mean_inter_token_latency_ms": 20.79022035217769, "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": 1955.0268999300897, "tpot_ms": 20.79022035217769, "ttft_ms": 1955.027399933897, "visible_content_truncated": false, "visible_generation_ms": 10623.802099958993 }, { "content_chunks": 512, "decode_tok_s": 47.282421977577826, "e2e_ms": 12583.28280004207, "effective_prefill_tok_s": 2054.7515552527616, "generation_ms": 10807.399000041187, "mean_inter_token_latency_ms": 21.14950880634283, "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": 1775.8838000008836, "tpot_ms": 21.14950880634283, "ttft_ms": 1775.8842000039294, "visible_content_truncated": false, "visible_generation_ms": 10807.398600038141 }, { "content_chunks": 512, "decode_tok_s": 44.82339206314004, "e2e_ms": 12590.445700101554, "effective_prefill_tok_s": 3066.844694834628, "generation_ms": 11400.297400075942, "mean_inter_token_latency_ms": 22.309779647898125, "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": 1190.1483000256121, "tpot_ms": 22.309779647898125, "ttft_ms": 1190.148700028658, "visible_content_truncated": false, "visible_generation_ms": 11400.297000072896 }, { "content_chunks": 512, "decode_tok_s": 41.995908562643415, "e2e_ms": 12599.042700021528, "effective_prefill_tok_s": 8467.251266479081, "generation_ms": 12167.852000100538, "mean_inter_token_latency_ms": 23.811843444423754, "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": 431.1906999209896, "tpot_ms": 23.811843444423754, "ttft_ms": 431.1911999247968, "visible_content_truncated": false, "visible_generation_ms": 12167.85150009673 }, { "content_chunks": 512, "decode_tok_s": 43.33987858762002, "e2e_ms": 12602.268800023012, "effective_prefill_tok_s": 4498.966420738841, "generation_ms": 11790.526800090447, "mean_inter_token_latency_ms": 23.07343796495195, "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": 811.7419999325648, "tpot_ms": 23.07343796495195, "ttft_ms": 811.7427999386564, "visible_content_truncated": false, "visible_generation_ms": 11790.526000084355 }, { "content_chunks": 512, "decode_tok_s": 46.36375342368831, "e2e_ms": 12602.763899951242, "effective_prefill_tok_s": 2310.2369299415623, "generation_ms": 11021.54079999309, "mean_inter_token_latency_ms": 21.568572994115637, "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": 1581.2230999581516, "tpot_ms": 21.568572994115637, "ttft_ms": 1581.2232999596745, "visible_content_truncated": false, "visible_generation_ms": 11021.540599991567 }, { "content_chunks": 512, "decode_tok_s": 46.36243383569609, "e2e_ms": 12602.694500004873, "effective_prefill_tok_s": 2310.796791479999, "generation_ms": 11021.854499937035, "mean_inter_token_latency_ms": 21.569186888330794, "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": 1580.8400000678375, "tpot_ms": 21.569186888330794, "ttft_ms": 1580.8405000716448, "visible_content_truncated": false, "visible_generation_ms": 11021.853999933228 }, { "content_chunks": 512, "decode_tok_s": 48.049774827383985, "e2e_ms": 12603.671899996698, "effective_prefill_tok_s": 1855.8900842142787, "generation_ms": 10634.80530004017, "mean_inter_token_latency_ms": 20.811752054873132, "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": 1968.8665999565274, "tpot_ms": 20.811752054873132, "ttft_ms": 1968.867199961096, "visible_content_truncated": false, "visible_generation_ms": 10634.804700035602 }, { "content_chunks": 512, "decode_tok_s": 48.05103723520087, "e2e_ms": 12603.232300025411, "effective_prefill_tok_s": 1855.533156171984, "generation_ms": 10634.525900008157, "mean_inter_token_latency_ms": 20.8112052837733, "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": 1968.706400017254, "tpot_ms": 20.8112052837733, "ttft_ms": 1968.7067000195384, "visible_content_truncated": false, "visible_generation_ms": 10634.525600005873 }, { "content_chunks": 512, "decode_tok_s": 47.23835210635074, "e2e_ms": 12604.031100054272, "effective_prefill_tok_s": 2045.283265490368, "generation_ms": 10817.481499980204, "mean_inter_token_latency_ms": 21.169239725988657, "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": 1786.5496000740677, "tpot_ms": 21.169239725988657, "ttft_ms": 1786.550100077875, "visible_content_truncated": false, "visible_generation_ms": 10817.480999976397 }, { "content_chunks": 512, "decode_tok_s": 44.77965526537698, "e2e_ms": 12605.643999995664, "effective_prefill_tok_s": 3060.596118710325, "generation_ms": 11411.432199995033, "mean_inter_token_latency_ms": 22.331569863003978, "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": 1194.2118000006303, "tpot_ms": 22.331569863003978, "ttft_ms": 1194.2122000036761, "visible_content_truncated": false, "visible_generation_ms": 11411.431799991988 }, { "content_chunks": 512, "decode_tok_s": 41.9374387627333, "e2e_ms": 12617.341700010002, "effective_prefill_tok_s": 8450.376638007341, "generation_ms": 12184.816600056365, "mean_inter_token_latency_ms": 23.845042270169014, "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": 432.5250999536365, "tpot_ms": 23.845042270169014, "ttft_ms": 432.5253999559209, "visible_content_truncated": false, "visible_generation_ms": 12184.81630005408 }, { "content_chunks": 512, "decode_tok_s": 43.28932229624109, "e2e_ms": 12617.88160004653, "effective_prefill_tok_s": 4492.462373343785, "generation_ms": 11804.296600050293, "mean_inter_token_latency_ms": 23.100384735910552, "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": 813.5849999962375, "tpot_ms": 23.100384735910552, "ttft_ms": 813.5856000008062, "visible_content_truncated": false, "visible_generation_ms": 11804.296000045724 }, { "content_chunks": 512, "decode_tok_s": 46.31229933231219, "e2e_ms": 12617.78209998738, "effective_prefill_tok_s": 2307.4551761809626, "generation_ms": 11033.785999985412, "mean_inter_token_latency_ms": 21.592536203493957, "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": 1583.9961000019684, "tpot_ms": 21.592536203493957, "ttft_ms": 1583.9966000057757, "visible_content_truncated": false, "visible_generation_ms": 11033.785499981605 }, { "content_chunks": 512, "decode_tok_s": 46.31417350987772, "e2e_ms": 12617.350200074725, "effective_prefill_tok_s": 2307.4339081216795, "generation_ms": 11033.339500078, "mean_inter_token_latency_ms": 21.591662426767122, "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": 1584.0106999967247, "tpot_ms": 21.591662426767122, "ttft_ms": 1584.0108999982476, "visible_content_truncated": false, "visible_generation_ms": 11033.339300076477 }, { "content_chunks": 512, "decode_tok_s": 48.00077432511693, "e2e_ms": 12617.815100005828, "effective_prefill_tok_s": 1853.3040150612355, "generation_ms": 10645.661599934101, "mean_inter_token_latency_ms": 20.83299726014501, "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": 1972.1535000717267, "tpot_ms": 20.83299726014501, "ttft_ms": 1972.1539000747725, "visible_content_truncated": false, "visible_generation_ms": 10645.661199931055 }, { "content_chunks": 512, "decode_tok_s": 48.00089381249531, "e2e_ms": 12617.357099894434, "effective_prefill_tok_s": 1854.216770990595, "generation_ms": 10645.635099965148, "mean_inter_token_latency_ms": 20.832945401105963, "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": 1971.7219999292865, "tpot_ms": 20.832945401105963, "ttft_ms": 1971.7226999346167, "visible_content_truncated": false, "visible_generation_ms": 10645.634399959818 }, { "content_chunks": 512, "decode_tok_s": 47.183004286755846, "e2e_ms": 12619.86989993602, "effective_prefill_tok_s": 2043.360363998356, "generation_ms": 10830.170899978839, "mean_inter_token_latency_ms": 21.19407221130888, "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": 1789.6989999571815, "tpot_ms": 21.19407221130888, "ttft_ms": 1789.6994999609888, "visible_content_truncated": false, "visible_generation_ms": 10830.170399975032 }, { "content_chunks": 512, "decode_tok_s": 44.72334501995974, "e2e_ms": 12621.69649999123, "effective_prefill_tok_s": 3058.7933870623965, "generation_ms": 11425.80009996891, "mean_inter_token_latency_ms": 22.359687084087888, "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": 1195.8964000223204, "tpot_ms": 22.359687084087888, "ttft_ms": 1195.8967000246048, "visible_content_truncated": false, "visible_generation_ms": 11425.799799966626 }, { "content_chunks": 512, "decode_tok_s": 41.892392674967226, "e2e_ms": 12630.69090002682, "effective_prefill_tok_s": 8454.794462842647, "generation_ms": 12197.918700054288, "mean_inter_token_latency_ms": 23.870682387581777, "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": 432.7721999725327, "tpot_ms": 23.870682387581777, "ttft_ms": 432.7729999786243, "visible_content_truncated": false, "visible_generation_ms": 12197.917900048196 }, { "content_chunks": 512, "decode_tok_s": 41.89041181381198, "e2e_ms": 13014.036499895155, "effective_prefill_tok_s": 4487.8185156224545, "generation_ms": 12198.49549990613, "mean_inter_token_latency_ms": 23.871811154415127, "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": 815.5409999890253, "tpot_ms": 23.871811154415127, "ttft_ms": 815.5416999943554, "visible_content_truncated": false, "visible_generation_ms": 12198.4947999008 }, { "content_chunks": 485, "decode_tok_s": 45.91988735029787, "e2e_ms": 12150.102899991907, "effective_prefill_tok_s": 2305.0813222797296, "generation_ms": 10561.872599995695, "mean_inter_token_latency_ms": 21.77705690720762, "output_token_source": "usage", "output_tokens": 486, "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": 1588.2302999962121, "tpot_ms": 21.77705690720762, "ttft_ms": 1588.2308000000194, "visible_content_truncated": false, "visible_generation_ms": 10561.872099991888 }, { "content_chunks": 512, "decode_tok_s": 44.72156371954676, "e2e_ms": 13014.32079996448, "effective_prefill_tok_s": 2305.3203847169348, "generation_ms": 11426.255199941806, "mean_inter_token_latency_ms": 22.360577690688466, "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": 1588.0656000226736, "tpot_ms": 22.360577690688466, "ttft_ms": 1588.0660000257194, "visible_content_truncated": false, "visible_generation_ms": 11426.25479993876 }, { "content_chunks": 512, "decode_tok_s": 46.29588734439582, "e2e_ms": 13015.592399984598, "effective_prefill_tok_s": 1851.4633916937403, "generation_ms": 11037.69749996718, "mean_inter_token_latency_ms": 21.60019080228411, "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": 1977.894900017418, "tpot_ms": 21.60019080228411, "ttft_ms": 1977.8951000189409, "visible_content_truncated": false, "visible_generation_ms": 11037.697299965657 }, { "content_chunks": 512, "decode_tok_s": 46.29655341580855, "e2e_ms": 13014.97819996439, "effective_prefill_tok_s": 1852.3954841206316, "generation_ms": 11037.538699922152, "mean_inter_token_latency_ms": 21.599880038986598, "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": 1977.4395000422373, "tpot_ms": 21.599880038986598, "ttft_ms": 1977.440100046806, "visible_content_truncated": false, "visible_generation_ms": 11037.538099917583 }, { "content_chunks": 512, "decode_tok_s": 46.29215929321879, "e2e_ms": 12833.064300008118, "effective_prefill_tok_s": 2042.3767826563674, "generation_ms": 11038.586399983615, "mean_inter_token_latency_ms": 21.60193033264895, "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": 1794.4779000245035, "tpot_ms": 21.60193033264895, "ttft_ms": 1794.4783000275493, "visible_content_truncated": false, "visible_generation_ms": 11038.585999980569 }, { "content_chunks": 512, "decode_tok_s": 44.72904449102429, "e2e_ms": 12623.901200015098, "effective_prefill_tok_s": 3055.2945793971962, "generation_ms": 11424.344199942425, "mean_inter_token_latency_ms": 22.35683796466228, "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": 1199.557000072673, "tpot_ms": 22.35683796466228, "ttft_ms": 1199.5575000764802, "visible_content_truncated": false, "visible_generation_ms": 11424.343699938618 }, { "content_chunks": 512, "decode_tok_s": 41.88276383297419, "e2e_ms": 12634.701199945994, "effective_prefill_tok_s": 8447.42892695537, "generation_ms": 12200.722999987192, "mean_inter_token_latency_ms": 23.876170254378067, "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": 433.97819995880127, "tpot_ms": 23.876170254378067, "ttft_ms": 433.9784999610856, "visible_content_truncated": false, "visible_generation_ms": 12200.722699984908 }, { "content_chunks": 512, "decode_tok_s": 41.882783400149286, "e2e_ms": 12635.3611999657, "effective_prefill_tok_s": 8436.791589195533, "generation_ms": 12200.71729994379, "mean_inter_token_latency_ms": 23.876159099694306, "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": 434.64390002191067, "tpot_ms": 23.876159099694306, "ttft_ms": 434.6444000257179, "visible_content_truncated": false, "visible_generation_ms": 12200.716799939983 }, { "content_chunks": 512, "decode_tok_s": 42.574351918770105, "e2e_ms": 12637.062800000422, "effective_prefill_tok_s": 5777.492773623463, "generation_ms": 12002.531500067562, "mean_inter_token_latency_ms": 23.488319960993273, "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": 634.5312999328598, "tpot_ms": 23.488319960993273, "ttft_ms": 634.5316999359056, "visible_content_truncated": false, "visible_generation_ms": 12002.531100064516 }, { "content_chunks": 512, "decode_tok_s": 44.67129224030871, "e2e_ms": 12636.678200098686, "effective_prefill_tok_s": 3062.0485262445986, "generation_ms": 11439.113900065422, "mean_inter_token_latency_ms": 22.38574148740787, "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": 1197.5643000332639, "tpot_ms": 22.38574148740787, "ttft_ms": 1197.564800037071, "visible_content_truncated": false, "visible_generation_ms": 11439.113400061615 }, { "content_chunks": 512, "decode_tok_s": 46.251940510688875, "e2e_ms": 12637.753500021063, "effective_prefill_tok_s": 2307.5446140833124, "generation_ms": 11048.185100080445, "mean_inter_token_latency_ms": 21.62071448156643, "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": 1589.5683999406174, "tpot_ms": 21.62071448156643, "ttft_ms": 1589.5687999436632, "visible_content_truncated": false, "visible_generation_ms": 11048.1847000774 }, { "content_chunks": 512, "decode_tok_s": 46.24992820393624, "e2e_ms": 12638.06599995587, "effective_prefill_tok_s": 2308.4179806555435, "generation_ms": 11048.665800015442, "mean_inter_token_latency_ms": 21.6216551859402, "output_token_source": "usage", "output_tokens": 512, "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": 1589.4001999404281, "tpot_ms": 21.6216551859402, "ttft_ms": 1589.400599943474, "visible_content_truncated": false, "visible_generation_ms": 11048.665400012396 }, { "content_chunks": 512, "decode_tok_s": 46.24941249256876, "e2e_ms": 12637.527199927717, "effective_prefill_tok_s": 2309.3798587768624, "generation_ms": 11048.78899990581, "mean_inter_token_latency_ms": 21.621896281616067, "output_token_source": "usage", "output_tokens": 512, "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": 1588.7382000219077, "tpot_ms": 21.621896281616067, "ttft_ms": 1588.7386000249535, "visible_content_truncated": false, "visible_generation_ms": 11048.788599902764 }, { "content_chunks": 512, "decode_tok_s": 44.69593004540696, "e2e_ms": 12636.939799995162, "effective_prefill_tok_s": 3047.839874680797, "generation_ms": 11432.808300014585, "mean_inter_token_latency_ms": 22.373401761280988, "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": 1204.1314999805763, "tpot_ms": 22.373401761280988, "ttft_ms": 1204.1318999836221, "visible_content_truncated": false, "visible_generation_ms": 11432.80790001154 }, { "content_chunks": 512, "decode_tok_s": 41.868902272643936, "e2e_ms": 12640.001700026914, "effective_prefill_tok_s": 8434.438609794599, "generation_ms": 12204.76230001077, "mean_inter_token_latency_ms": 23.8840749510974, "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": 435.2394000161439, "tpot_ms": 23.8840749510974, "ttft_ms": 435.2398000191897, "visible_content_truncated": false, "visible_generation_ms": 12204.761900007725 }, { "content_chunks": 512, "decode_tok_s": 41.85907197307979, "e2e_ms": 12642.901999992318, "effective_prefill_tok_s": 8436.075248407089, "generation_ms": 12207.628499949351, "mean_inter_token_latency_ms": 23.88968395293415, "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": 435.2735000429675, "tpot_ms": 23.88968395293415, "ttft_ms": 435.27460005134344, "visible_content_truncated": false, "visible_generation_ms": 12207.627399940975 }, { "content_chunks": 512, "decode_tok_s": 42.556968936332815, "e2e_ms": 12642.527800053358, "effective_prefill_tok_s": 5781.824004646095, "generation_ms": 12007.434100029059, "mean_inter_token_latency_ms": 23.497914090076435, "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": 635.0937000242993, "tpot_ms": 23.497914090076435, "ttft_ms": 635.0944000296295, "visible_content_truncated": false, "visible_generation_ms": 12007.433400023729 }, { "content_chunks": 512, "decode_tok_s": 44.64917240478197, "e2e_ms": 12642.616299912333, "effective_prefill_tok_s": 3065.5299605847035, "generation_ms": 11444.780999911018, "mean_inter_token_latency_ms": 22.396831702369898, "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": 1197.8353000013158, "tpot_ms": 22.396831702369898, "ttft_ms": 1197.8357998887077, "visible_content_truncated": false, "visible_generation_ms": 11444.780500023626 }, { "content_chunks": 512, "decode_tok_s": 46.22821380100371, "e2e_ms": 12642.173299915157, "effective_prefill_tok_s": 2311.8800477291256, "generation_ms": 11053.85559995193, "mean_inter_token_latency_ms": 21.631811350199474, "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": 1588.317699963227, "tpot_ms": 21.631811350199474, "ttft_ms": 1588.3184999693185, "visible_content_truncated": false, "visible_generation_ms": 11053.854799945839 }, { "content_chunks": 512, "decode_tok_s": 46.22733892409828, "e2e_ms": 12642.021500039846, "effective_prefill_tok_s": 2312.405621627254, "generation_ms": 11054.064800031483, "mean_inter_token_latency_ms": 21.632220743701534, "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": 1587.9567000083625, "tpot_ms": 21.632220743701534, "ttft_ms": 1587.9570000106469, "visible_content_truncated": false, "visible_generation_ms": 11054.064500029199 }, { "content_chunks": 512, "decode_tok_s": 46.23131000896272, "e2e_ms": 12642.292200005613, "effective_prefill_tok_s": 2310.630113004763, "generation_ms": 11053.115300019272, "mean_inter_token_latency_ms": 21.63036262234691, "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": 1589.1768999863416, "tpot_ms": 21.63036262234691, "ttft_ms": 1589.1772999893874, "visible_content_truncated": false, "visible_generation_ms": 11053.114900016226 }, { "content_chunks": 512, "decode_tok_s": 44.66962090475892, "e2e_ms": 12643.31359998323, "effective_prefill_tok_s": 3052.0737447548777, "generation_ms": 11439.541900064796, "mean_inter_token_latency_ms": 22.386579060792165, "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": 1203.771699918434, "tpot_ms": 22.386579060792165, "ttft_ms": 1203.7720999214798, "visible_content_truncated": false, "visible_generation_ms": 11439.54150006175 }, { "content_chunks": 512, "decode_tok_s": 41.84908760387273, "e2e_ms": 12645.580200012773, "effective_prefill_tok_s": 8445.215971242931, "generation_ms": 12210.541000007652, "mean_inter_token_latency_ms": 23.89538356165881, "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": 435.03920000512153, "tpot_ms": 23.89538356165881, "ttft_ms": 435.0395000074059, "visible_content_truncated": false, "visible_generation_ms": 12210.540700005367 }, { "content_chunks": 512, "decode_tok_s": 41.84816431197119, "e2e_ms": 12646.48690004833, "effective_prefill_tok_s": 8435.15773580768, "generation_ms": 12210.81039996352, "mean_inter_token_latency_ms": 23.895910763138005, "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": 435.67650008480996, "tpot_ms": 23.895910763138005, "ttft_ms": 435.67690008785576, "visible_content_truncated": false, "visible_generation_ms": 12210.809999960475 }, { "content_chunks": 512, "decode_tok_s": 42.54286053719444, "e2e_ms": 12646.382800070569, "effective_prefill_tok_s": 5789.27997336885, "generation_ms": 12011.416100081988, "mean_inter_token_latency_ms": 23.5057066537808, "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": 634.9666999885812, "tpot_ms": 23.5057066537808, "ttft_ms": 634.9669999908656, "visible_content_truncated": false, "visible_generation_ms": 12011.415800079703 }, { "content_chunks": 512, "decode_tok_s": 44.63605628990572, "e2e_ms": 12646.72070008237, "effective_prefill_tok_s": 3067.8053392992856, "generation_ms": 11448.144000023603, "mean_inter_token_latency_ms": 22.403412915897462, "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": 1198.5767000587657, "tpot_ms": 22.403412915897462, "ttft_ms": 1198.5774000640959, "visible_content_truncated": false, "visible_generation_ms": 11448.143300018273 }, { "content_chunks": 512, "decode_tok_s": 46.21114714124146, "e2e_ms": 12647.63560006395, "effective_prefill_tok_s": 2313.0185263008443, "generation_ms": 11057.938000070862, "mean_inter_token_latency_ms": 21.639800391528105, "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": 1589.6975999930874, "tpot_ms": 21.639800391528105, "ttft_ms": 1589.6979999961331, "visible_content_truncated": false, "visible_generation_ms": 11057.937600067817 }, { "content_chunks": 512, "decode_tok_s": 46.21311720726414, "e2e_ms": 12646.912000025623, "effective_prefill_tok_s": 2314.0146870601516, "generation_ms": 11057.46659997385, "mean_inter_token_latency_ms": 21.63887788644589, "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": 1589.4454000517726, "tpot_ms": 21.63887788644589, "ttft_ms": 1589.4456000532955, "visible_content_truncated": false, "visible_generation_ms": 11057.466399972327 }, { "content_chunks": 512, "decode_tok_s": 46.21377713707068, "e2e_ms": 12648.372700088657, "effective_prefill_tok_s": 2312.289134764815, "generation_ms": 11057.30870005209, "mean_inter_token_latency_ms": 21.638568884642055, "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": 1591.0640000365674, "tpot_ms": 21.638568884642055, "ttft_ms": 1591.0643000388518, "visible_content_truncated": false, "visible_generation_ms": 11057.308400049806 }, { "content_chunks": 512, "decode_tok_s": 44.65650175185955, "e2e_ms": 12648.235999979079, "effective_prefill_tok_s": 3053.09717613493, "generation_ms": 11442.902599927038, "mean_inter_token_latency_ms": 22.393155772851348, "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": 1205.3334000520408, "tpot_ms": 22.393155772851348, "ttft_ms": 1205.333900055848, "visible_content_truncated": false, "visible_generation_ms": 11442.90209992323 }, { "content_chunks": 512, "decode_tok_s": 41.829195465489235, "e2e_ms": 12652.281699934974, "effective_prefill_tok_s": 8439.35284782357, "generation_ms": 12216.34779998567, "mean_inter_token_latency_ms": 23.90674716239857, "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": 435.93389994930476, "tpot_ms": 23.90674716239857, "ttft_ms": 435.93429995235056, "visible_content_truncated": false, "visible_generation_ms": 12216.347399982624 }, { "content_chunks": 512, "decode_tok_s": 41.82529380000378, "e2e_ms": 12671.941899927333, "effective_prefill_tok_s": 8097.620333911905, "generation_ms": 12217.487399931997, "mean_inter_token_latency_ms": 23.908977299279837, "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": 454.45449999533594, "tpot_ms": 23.908977299279837, "ttft_ms": 454.4552000006661, "visible_content_truncated": false, "visible_generation_ms": 12217.486699926667 }, { "content_chunks": 512, "decode_tok_s": 41.214421066327986, "e2e_ms": 13035.439100000076, "effective_prefill_tok_s": 5779.863057439891, "generation_ms": 12398.572799982503, "mean_inter_token_latency_ms": 24.263351859065565, "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": 636.8663000175729, "tpot_ms": 24.263351859065565, "ttft_ms": 636.8669000221416, "visible_content_truncated": false, "visible_generation_ms": 12398.572199977934 }, { "content_chunks": 512, "decode_tok_s": 43.17601006378813, "e2e_ms": 13035.544399986975, "effective_prefill_tok_s": 3067.64745844475, "generation_ms": 11835.276099969633, "mean_inter_token_latency_ms": 23.1610099803711, "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": 1200.268300017342, "tpot_ms": 23.1610099803711, "ttft_ms": 1200.2687000203878, "visible_content_truncated": false, "visible_generation_ms": 11835.275699966587 }, { "content_chunks": 497, "decode_tok_s": 46.889103731957604, "e2e_ms": 12191.418299917132, "effective_prefill_tok_s": 2313.5282418105853, "generation_ms": 10599.477499956265, "mean_inter_token_latency_ms": 21.326916498905966, "output_token_source": "usage", "output_tokens": 498, "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": 1591.9407999608666, "tpot_ms": 21.326916498905966, "ttft_ms": 1591.941099963151, "visible_content_truncated": false, "visible_generation_ms": 10599.47719995398 }, { "content_chunks": 512, "decode_tok_s": 44.65172437030989, "e2e_ms": 13035.552800050937, "effective_prefill_tok_s": 2314.905142617052, "generation_ms": 11444.126900052652, "mean_inter_token_latency_ms": 22.395551663508126, "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": 1591.4258999982849, "tpot_ms": 22.395551663508126, "ttft_ms": 1591.4267000043765, "visible_content_truncated": false, "visible_generation_ms": 11444.12610004656 }, { "content_chunks": 512, "decode_tok_s": 45.44848740819624, "e2e_ms": 12835.985800018534, "effective_prefill_tok_s": 2313.9901962732993, "generation_ms": 11243.498499970883, "mean_inter_token_latency_ms": 22.002932485265916, "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": 1592.487300047651, "tpot_ms": 22.002932485265916, "ttft_ms": 1592.4877000506967, "visible_content_truncated": false, "visible_generation_ms": 11243.498099967837 }, { "content_chunks": 512, "decode_tok_s": 46.188867091283264, "e2e_ms": 12270.472800009884, "effective_prefill_tok_s": 3053.3445635475946, "generation_ms": 11063.271999941207, "mean_inter_token_latency_ms": 21.650238747438763, "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": 1207.2008000686765, "tpot_ms": 21.650238747438763, "ttft_ms": 1207.2013000724837, "visible_content_truncated": false, "visible_generation_ms": 11063.2714999374 }, { "content_chunks": 512, "decode_tok_s": 48.01028513748257, "e2e_ms": 11080.753599992022, "effective_prefill_tok_s": 8430.906705753838, "generation_ms": 10643.55269994121, "mean_inter_token_latency_ms": 20.828870254288084, "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": 437.20090005081147, "tpot_ms": 20.828870254288084, "ttft_ms": 437.2014000546187, "visible_content_truncated": false, "visible_generation_ms": 10643.552199937403 }, { "content_chunks": 512, "decode_tok_s": 49.94910079109986, "e2e_ms": 10666.873799986206, "effective_prefill_tok_s": 8447.521121237189, "generation_ms": 10230.414399993606, "mean_inter_token_latency_ms": 20.020380430515864, "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": 436.4593999926001, "tpot_ms": 20.020380430515864, "ttft_ms": 436.4597999956459, "visible_content_truncated": false, "visible_generation_ms": 10230.41399999056 }, { "content_chunks": 512, "decode_tok_s": 51.9270715476379, "e2e_ms": 10277.705100015737, "effective_prefill_tok_s": 8442.020436982579, "generation_ms": 9840.724400011823, "mean_inter_token_latency_ms": 19.257777690825485, "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": 436.98070000391454, "tpot_ms": 19.257777690825485, "ttft_ms": 436.98110000696033, "visible_content_truncated": false, "visible_generation_ms": 9840.724000008777 }, { "content_chunks": 512, "decode_tok_s": 54.17163288593884, "e2e_ms": 9870.348499971442, "effective_prefill_tok_s": 8434.560249812648, "generation_ms": 9432.981299934909, "mean_inter_token_latency_ms": 18.459845988130937, "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": 437.3672000365332, "tpot_ms": 18.459845988130937, "ttft_ms": 437.3678000411019, "visible_content_truncated": false, "visible_generation_ms": 9432.98069993034 } ], "requests": 100, "response_words": 256, "run_id": "benchmark-20260905T051645Z", "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": "2026-09-05T05:16:45Z", "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": 161922.2221999662 }