{ "dataset": "search-benchmark-lanes", "title": "OpenRouter web search benchmark lanes", "description": "One row per lane (suite, model, search engine, surface, turn budget, reasoning effort) on the OpenRouter web search benchmark pages (BrowseComp, HLE, DeepSearchQA, WideSearch), pooled across the independent runs the page displays.", "schema_version": "v1", "snapshot_date": "2026-09-22", "retrieved_at": "2026-09-22T01:45:08.571Z", "update_cadence": "daily", "license": { "name": "CC BY 4.0", "url": "https://creativecommons.org/licenses/by/4.0/" }, "attribution": "Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.", "harness": { "name": "OpenRouter benchmark harness", "url": "https://github.com/OpenRouterTeam/benchmark-harness" }, "source_pages": [ "https://openrouter.ai/benchmarks/browsecomp", "https://openrouter.ai/benchmarks/hle", "https://openrouter.ai/benchmarks/deepsearchqa", "https://openrouter.ai/benchmarks/widesearch" ], "methodology": "Runs pair a model with a web search engine through the OpenRouter server-tool or plugin surface. A run contributes only when it succeeded, is visible on openrouter.ai, has no epoch marker, ran through a public search surface, answered at least 45 questions, and is the latest run of its resume lineage; within a lane only runs sharing the most recent configuration are pooled. Score is pooled correct answers over pooled questions, which is the value the suite pages plot.", "readme_url": "https://openrouter.ai/api/v1/datasets/exports/README.md", "columns": [ { "name": "benchmark", "type": "string", "nullable": false, "description": "Search suite identifier." }, { "name": "benchmark_page_url", "type": "string", "nullable": false, "description": "openrouter.ai suite page that displays this lane." }, { "name": "model_permaslug", "type": "string", "nullable": false, "description": "Immutable OpenRouter model version identifier." }, { "name": "model_name", "type": "string", "nullable": true, "description": "Display name, null when the version is no longer listed." }, { "name": "search_engine", "type": "string", "nullable": false, "description": "Web search engine used." }, { "name": "surface", "type": "string", "nullable": false, "description": "How search was invoked: server-tool or plugin." }, { "name": "max_agent_turns", "type": "integer", "nullable": true, "description": "Agent turn budget, null when the lane ran with the server default." }, { "name": "reasoning_effort", "type": "string", "nullable": true, "description": "Reasoning effort requested from the model, null when unset." }, { "name": "score", "type": "number", "nullable": false, "description": "Pooled correct answers divided by pooled questions, 0 to 1." }, { "name": "item_f1", "type": "number", "nullable": false, "description": "Sample-weighted partial-credit metric (item F1 for WideSearch, accuracy otherwise)." }, { "name": "runs", "type": "integer", "nullable": false, "description": "Independent runs pooled into the lane." }, { "name": "total_questions", "type": "integer", "nullable": false, "description": "Questions pooled across the runs (sample size)." }, { "name": "total_correct", "type": "integer", "nullable": false, "description": "Correct answers pooled across the runs." }, { "name": "cost_per_task_usd", "type": "number", "nullable": true, "description": "Mean inference and search spend per question in USD." }, { "name": "latency_mean_ms", "type": "number", "nullable": true, "description": "Sample-weighted mean generation time per question in milliseconds, null when no run recorded timing." }, { "name": "last_run_at", "type": "string", "nullable": false, "description": "ISO 8601 UTC timestamp of the most recent pooled run." } ], "row_count": 134, "rows": [ { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.8900523560209425, "item_f1": 0.890052, "runs": 1, "total_questions": 573, "total_correct": 510, "cost_per_task_usd": 0.9906557085514834, "latency_mean_ms": 112662.38219895288, "last_run_at": "2026-08-11T00:51:14.223Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.8634146341463415, "item_f1": 0.8634146097560976, "runs": 2, "total_questions": 820, "total_correct": 708, "cost_per_task_usd": 1.6815852951219512, "latency_mean_ms": 165960.47073170732, "last_run_at": "2026-08-17T13:04:36.414Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.8243670886075949, "item_f1": 0.8243670000000001, "runs": 1, "total_questions": 632, "total_correct": 521, "cost_per_task_usd": 0.5028527816455696, "latency_mean_ms": 115870.25, "last_run_at": "2026-08-11T00:51:21.328Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.8218694885361552, "item_f1": 0.8218690000000001, "runs": 1, "total_questions": 567, "total_correct": 466, "cost_per_task_usd": 1.2943719012345678, "latency_mean_ms": 141607.0405643739, "last_run_at": "2026-08-10T23:33:31.668Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7792946530147895, "item_f1": 0.7792947622298065, "runs": 2, "total_questions": 879, "total_correct": 685, "cost_per_task_usd": 1.0722636029579067, "latency_mean_ms": 148158.76791808874, "last_run_at": "2026-08-17T13:04:37.736Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7784045124899275, "item_f1": 0.778405, "runs": 1, "total_questions": 1241, "total_correct": 966, "cost_per_task_usd": 0.5407230032232071, "latency_mean_ms": 144278.3521353747, "last_run_at": "2026-08-10T22:20:35.997Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7777777777777778, "item_f1": 0.777778, "runs": 1, "total_questions": 99, "total_correct": 77, "cost_per_task_usd": 0.6815806464646464, "latency_mean_ms": 171980.27272727274, "last_run_at": "2026-08-11T11:09:32.883Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.77, "item_f1": 0.77, "runs": 1, "total_questions": 100, "total_correct": 77, "cost_per_task_usd": 0.07649624000000001, "latency_mean_ms": 135422.98, "last_run_at": "2026-08-11T10:55:08.930Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.74, "item_f1": 0.74, "runs": 1, "total_questions": 100, "total_correct": 74, "cost_per_task_usd": 0.09858238, "latency_mean_ms": 111189.6, "last_run_at": "2026-08-11T05:36:23.702Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6907216494845361, "item_f1": 0.6907215463917525, "runs": 2, "total_questions": 194, "total_correct": 134, "cost_per_task_usd": 0.059997422680412374, "latency_mean_ms": 158496.7525773196, "last_run_at": "2026-08-18T02:36:37.807Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.6836734693877551, "item_f1": 0.683673, "runs": 1, "total_questions": 98, "total_correct": 67, "cost_per_task_usd": 0.13989858163265306, "latency_mean_ms": 136037.64285714287, "last_run_at": "2026-08-11T05:35:50.346Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6736842105263158, "item_f1": 0.6736840000000001, "runs": 1, "total_questions": 95, "total_correct": 64, "cost_per_task_usd": 0.1196647052631579, "latency_mean_ms": 176100.3052631579, "last_run_at": "2026-08-11T10:51:47.598Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6645367412140575, "item_f1": 0.6645369999999999, "runs": 1, "total_questions": 626, "total_correct": 416, "cost_per_task_usd": 0.5111501916932908, "latency_mean_ms": 95484.04153354633, "last_run_at": "2026-08-11T00:55:16.432Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6518987341772152, "item_f1": 0.651899, "runs": 1, "total_questions": 632, "total_correct": 412, "cost_per_task_usd": 0.28770013607594935, "latency_mean_ms": 102526.72151898734, "last_run_at": "2026-08-11T00:50:53.218Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6020293122886133, "item_f1": 0.6020292694475761, "runs": 2, "total_questions": 887, "total_correct": 534, "cost_per_task_usd": 0.4730216178128523, "latency_mean_ms": 118330.07891770011, "last_run_at": "2026-08-17T07:48:29.613Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6009501187648456, "item_f1": 0.60095, "runs": 1, "total_questions": 1263, "total_correct": 759, "cost_per_task_usd": 0.3016604172604909, "latency_mean_ms": 126625.86064924783, "last_run_at": "2026-08-10T22:20:36.085Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.59, "item_f1": 0.59, "runs": 1, "total_questions": 100, "total_correct": 59, "cost_per_task_usd": 0.34904374, "latency_mean_ms": 153563.16, "last_run_at": "2026-08-11T11:09:29.955Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5778732545649838, "item_f1": 0.577873365198711, "runs": 2, "total_questions": 931, "total_correct": 538, "cost_per_task_usd": 0.8407766444683137, "latency_mean_ms": 126304.41675617616, "last_run_at": "2026-08-17T05:48:06.984Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.57, "item_f1": 0.57, "runs": 1, "total_questions": 100, "total_correct": 57, "cost_per_task_usd": 0.036947299999999995, "latency_mean_ms": 88669.69, "last_run_at": "2026-08-11T05:15:28.464Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.56, "item_f1": 0.56, "runs": 2, "total_questions": 200, "total_correct": 112, "cost_per_task_usd": 0.07293105, "latency_mean_ms": 131584.81, "last_run_at": "2026-08-18T02:39:26.886Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5248, "item_f1": 0.5248, "runs": 1, "total_questions": 625, "total_correct": 328, "cost_per_task_usd": 0.5376877936, "latency_mean_ms": 102583.5984, "last_run_at": "2026-08-10T23:34:40.789Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.48, "item_f1": 0.48, "runs": 1, "total_questions": 100, "total_correct": 48, "cost_per_task_usd": 0.04922616, "latency_mean_ms": 104743.2, "last_run_at": "2026-08-11T05:15:23.073Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.4672, "item_f1": 0.4672, "runs": 1, "total_questions": 1250, "total_correct": 584, "cost_per_task_usd": 0.19860231679999998, "latency_mean_ms": 146687.088, "last_run_at": "2026-08-10T22:20:35.914Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.4634920634920635, "item_f1": 0.46349199999999996, "runs": 1, "total_questions": 630, "total_correct": 292, "cost_per_task_usd": 0.19563253333333333, "latency_mean_ms": 125923.51587301587, "last_run_at": "2026-08-11T00:50:53.051Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.4365079365079365, "item_f1": 0.4365081201814059, "runs": 2, "total_questions": 882, "total_correct": 385, "cost_per_task_usd": 0.20920909183673472, "latency_mean_ms": 132046.35941043083, "last_run_at": "2026-08-17T07:48:29.589Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.425, "item_f1": 0.425, "runs": 2, "total_questions": 200, "total_correct": 85, "cost_per_task_usd": 0.028510745, "latency_mean_ms": 114349.49, "last_run_at": "2026-08-18T02:37:59.229Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.41836734693877553, "item_f1": 0.418367, "runs": 1, "total_questions": 98, "total_correct": 41, "cost_per_task_usd": 0.2556308979591837, "latency_mean_ms": 177219.01020408163, "last_run_at": "2026-08-11T11:09:29.755Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.358161648177496, "item_f1": 0.35816200000000004, "runs": 1, "total_questions": 631, "total_correct": 226, "cost_per_task_usd": 0.13635232171156894, "latency_mean_ms": 47843.17432646592, "last_run_at": "2026-08-11T00:54:00.184Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.3522727272727273, "item_f1": 0.352273, "runs": 1, "total_questions": 88, "total_correct": 31, "cost_per_task_usd": 0.006993340909090909, "latency_mean_ms": 122042.03409090909, "last_run_at": "2026-08-18T02:39:27.062Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.3368421052631579, "item_f1": 0.33684200000000003, "runs": 1, "total_questions": 95, "total_correct": 32, "cost_per_task_usd": 0.01695815789473684, "latency_mean_ms": 140292.84210526315, "last_run_at": "2026-08-11T04:58:13.390Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.32989690721649484, "item_f1": 0.329897, "runs": 1, "total_questions": 97, "total_correct": 32, "cost_per_task_usd": 0.01931864948453608, "latency_mean_ms": 147455.51546391752, "last_run_at": "2026-08-11T04:57:42.950Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.3080168776371308, "item_f1": 0.30801712236286916, "runs": 2, "total_questions": 948, "total_correct": 292, "cost_per_task_usd": 0.1839051687763713, "latency_mean_ms": 58512.93565400844, "last_run_at": "2026-08-17T05:49:10.322Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.29430379746835444, "item_f1": 0.294304, "runs": 1, "total_questions": 632, "total_correct": 186, "cost_per_task_usd": 0.14004478481012658, "latency_mean_ms": 50013.446202531646, "last_run_at": "2026-08-10T23:33:31.756Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.23737373737373738, "item_f1": 0.23737375757575757, "runs": 2, "total_questions": 198, "total_correct": 47, "cost_per_task_usd": 0.01764588383838384, "latency_mean_ms": 133619.22222222222, "last_run_at": "2026-08-18T02:37:59.313Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.2, "item_f1": 0.2, "runs": 1, "total_questions": 85, "total_correct": 17, "cost_per_task_usd": 0.004914, "latency_mean_ms": 126236.55294117647, "last_run_at": "2026-08-18T02:36:38.570Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "plugin", "max_agent_turns": null, "reasoning_effort": "medium", "score": 0.12, "item_f1": 0.12, "runs": 1, "total_questions": 100, "total_correct": 12, "cost_per_task_usd": 0.43839346, "latency_mean_ms": 200985.09, "last_run_at": "2026-07-28T23:09:14.398Z" }, { "benchmark": "search_browsecomp", "benchmark_page_url": "https://openrouter.ai/benchmarks/browsecomp", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "plugin", "max_agent_turns": null, "reasoning_effort": "medium", "score": 0.11, "item_f1": 0.11, "runs": 1, "total_questions": 100, "total_correct": 11, "cost_per_task_usd": 0.42078111999999995, "latency_mean_ms": 177250.96, "last_run_at": "2026-07-28T23:11:58.519Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7704697986577181, "item_f1": 0.8797722416107383, "runs": 2, "total_questions": 745, "total_correct": 574, "cost_per_task_usd": 3.3245481422818797, "latency_mean_ms": 141832.2053691275, "last_run_at": "2026-08-14T21:13:31.975Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7651006711409396, "item_f1": 0.8702580000000001, "runs": 1, "total_questions": 447, "total_correct": 342, "cost_per_task_usd": 1.693907908277405, "latency_mean_ms": 113879.8769574944, "last_run_at": "2026-08-11T00:51:38.656Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.75, "item_f1": 0.856438, "runs": 1, "total_questions": 100, "total_correct": 75, "cost_per_task_usd": 0.57360524, "latency_mean_ms": 121172.67, "last_run_at": "2026-08-11T11:09:32.872Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7355555555555555, "item_f1": 0.873839, "runs": 2, "total_questions": 900, "total_correct": 662, "cost_per_task_usd": 1.167155038888889, "latency_mean_ms": 124780.71777777778, "last_run_at": "2026-08-14T20:37:57.151Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7355555555555555, "item_f1": 0.8660530000000001, "runs": 1, "total_questions": 450, "total_correct": 331, "cost_per_task_usd": 0.5691047644444445, "latency_mean_ms": 103340.41777777778, "last_run_at": "2026-08-11T00:53:59.790Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.73, "item_f1": 0.869804, "runs": 1, "total_questions": 100, "total_correct": 73, "cost_per_task_usd": 0.10303963, "latency_mean_ms": 96187.19, "last_run_at": "2026-08-11T05:36:00.481Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.704954954954955, "item_f1": 0.817238, "runs": 1, "total_questions": 444, "total_correct": 313, "cost_per_task_usd": 1.8020249391891892, "latency_mean_ms": 133394.7972972973, "last_run_at": "2026-08-10T23:34:06.938Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.69, "item_f1": 0.8151229999999999, "runs": 1, "total_questions": 100, "total_correct": 69, "cost_per_task_usd": 0.06348704, "latency_mean_ms": 78573.34, "last_run_at": "2026-08-11T10:52:17.086Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.69, "item_f1": 0.838773, "runs": 1, "total_questions": 900, "total_correct": 621, "cost_per_task_usd": 0.6383792044444445, "latency_mean_ms": 134197.76, "last_run_at": "2026-08-10T22:25:47.996Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.67, "item_f1": 0.823757, "runs": 1, "total_questions": 100, "total_correct": 67, "cost_per_task_usd": 0.14365725, "latency_mean_ms": 110946.58, "last_run_at": "2026-08-11T05:36:20.652Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.67, "item_f1": 0.81963, "runs": 1, "total_questions": 100, "total_correct": 67, "cost_per_task_usd": 0.25432224000000003, "latency_mean_ms": 64400.9, "last_run_at": "2026-08-11T11:09:32.871Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6466666666666666, "item_f1": 0.7870809999999999, "runs": 1, "total_questions": 450, "total_correct": 291, "cost_per_task_usd": 0.43797520888888886, "latency_mean_ms": 47579.07777777778, "last_run_at": "2026-08-11T00:55:44.528Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.64, "item_f1": 0.7769865, "runs": 2, "total_questions": 200, "total_correct": 128, "cost_per_task_usd": 0.059285875, "latency_mean_ms": 127017.09, "last_run_at": "2026-08-18T02:36:37.557Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.63, "item_f1": 0.829651, "runs": 2, "total_questions": 200, "total_correct": 126, "cost_per_task_usd": 0.07594056, "latency_mean_ms": 107434.355, "last_run_at": "2026-08-18T02:36:39.865Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6, "item_f1": 0.758802, "runs": 1, "total_questions": 100, "total_correct": 60, "cost_per_task_usd": 0.09611798, "latency_mean_ms": 91703.77, "last_run_at": "2026-08-11T10:57:34.059Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.6, "item_f1": 0.798222, "runs": 1, "total_questions": 100, "total_correct": 60, "cost_per_task_usd": 0.03247644, "latency_mean_ms": 64394.42, "last_run_at": "2026-08-11T05:15:25.346Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5977777777777777, "item_f1": 0.780041, "runs": 1, "total_questions": 450, "total_correct": 269, "cost_per_task_usd": 0.24649364444444444, "latency_mean_ms": 64465.47111111111, "last_run_at": "2026-08-11T00:53:42.779Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5944444444444444, "item_f1": 0.7713620000000001, "runs": 2, "total_questions": 900, "total_correct": 535, "cost_per_task_usd": 0.39104313444444444, "latency_mean_ms": 69591.83444444444, "last_run_at": "2026-08-14T20:37:57.063Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.58, "item_f1": 0.7173733999999999, "runs": 2, "total_questions": 750, "total_correct": 435, "cost_per_task_usd": 0.7350843586666667, "latency_mean_ms": 57044.516, "last_run_at": "2026-08-14T21:13:31.978Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5788888888888889, "item_f1": 0.7623650000000001, "runs": 1, "total_questions": 900, "total_correct": 521, "cost_per_task_usd": 0.2413582211111111, "latency_mean_ms": 82138.6888888889, "last_run_at": "2026-08-10T22:20:36.170Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.56, "item_f1": 0.74012, "runs": 1, "total_questions": 100, "total_correct": 56, "cost_per_task_usd": 0.02889355, "latency_mean_ms": 89422.54, "last_run_at": "2026-08-11T05:38:01.864Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.5577777777777778, "item_f1": 0.7052799999999999, "runs": 1, "total_questions": 450, "total_correct": 251, "cost_per_task_usd": 0.40950537555555555, "latency_mean_ms": 50486.07111111111, "last_run_at": "2026-08-10T23:35:58.072Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.55, "item_f1": 0.72724, "runs": 1, "total_questions": 100, "total_correct": 55, "cost_per_task_usd": 0.04319829, "latency_mean_ms": 70495.19, "last_run_at": "2026-08-11T05:17:54.251Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.53, "item_f1": 0.699294, "runs": 1, "total_questions": 100, "total_correct": 53, "cost_per_task_usd": 0.03898867, "latency_mean_ms": 101284.03, "last_run_at": "2026-08-11T05:37:03.485Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.515, "item_f1": 0.654662, "runs": 2, "total_questions": 200, "total_correct": 103, "cost_per_task_usd": 0.02256679, "latency_mean_ms": 107478.415, "last_run_at": "2026-08-18T02:37:59.137Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.495, "item_f1": 0.7002665, "runs": 2, "total_questions": 200, "total_correct": 99, "cost_per_task_usd": 0.028381314999999997, "latency_mean_ms": 69678.87, "last_run_at": "2026-08-18T02:36:38.057Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.45111111111111113, "item_f1": 0.662417, "runs": 1, "total_questions": 450, "total_correct": 203, "cost_per_task_usd": 0.12125408666666666, "latency_mean_ms": 58866.93777777778, "last_run_at": "2026-08-11T00:53:19.967Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.45, "item_f1": 0.652112, "runs": 1, "total_questions": 100, "total_correct": 45, "cost_per_task_usd": 0.14555545, "latency_mean_ms": 65981.47, "last_run_at": "2026-08-11T11:09:32.876Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.4266666666666667, "item_f1": 0.642664, "runs": 2, "total_questions": 900, "total_correct": 384, "cost_per_task_usd": 0.13944749, "latency_mean_ms": 56774.72222222222, "last_run_at": "2026-08-14T20:37:57.239Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.42, "item_f1": 0.602781, "runs": 1, "total_questions": 100, "total_correct": 42, "cost_per_task_usd": 0.01246064, "latency_mean_ms": 65688.2, "last_run_at": "2026-08-11T04:58:12.886Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.4166666666666667, "item_f1": 0.633481, "runs": 1, "total_questions": 900, "total_correct": 375, "cost_per_task_usd": 0.12751975444444444, "latency_mean_ms": 88179.57666666666, "last_run_at": "2026-08-10T22:23:47.583Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.3888888888888889, "item_f1": 0.565146, "runs": 1, "total_questions": 450, "total_correct": 175, "cost_per_task_usd": 0.09267309777777778, "latency_mean_ms": 20353.69111111111, "last_run_at": "2026-08-11T00:51:14.319Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.38, "item_f1": 0.5895199999999999, "runs": 1, "total_questions": 100, "total_correct": 38, "cost_per_task_usd": 0.0098914, "latency_mean_ms": 54892.89, "last_run_at": "2026-08-11T04:57:50.945Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.348, "item_f1": 0.5041728, "runs": 2, "total_questions": 750, "total_correct": 261, "cost_per_task_usd": 0.11558385866666666, "latency_mean_ms": 24094.12, "last_run_at": "2026-08-14T21:13:31.978Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.3333333333333333, "item_f1": 0.515215, "runs": 1, "total_questions": 450, "total_correct": 150, "cost_per_task_usd": 0.08556107777777777, "latency_mean_ms": 19247.51777777778, "last_run_at": "2026-08-10T23:35:57.985Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.275, "item_f1": 0.4723655, "runs": 2, "total_questions": 200, "total_correct": 55, "cost_per_task_usd": 0.01123002, "latency_mean_ms": 58866.955, "last_run_at": "2026-08-18T02:36:37.892Z" }, { "benchmark": "search_dsqa", "benchmark_page_url": "https://openrouter.ai/benchmarks/deepsearchqa", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.22, "item_f1": 0.373051, "runs": 1, "total_questions": 100, "total_correct": 22, "cost_per_task_usd": 0.0033189200000000004, "latency_mean_ms": 67209.76, "last_run_at": "2026-08-18T02:36:38.485Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7738095238095238, "item_f1": 0.77381, "runs": 1, "total_questions": 84, "total_correct": 65, "cost_per_task_usd": 0.4575818571428571, "latency_mean_ms": 82862.51190476191, "last_run_at": "2026-08-11T11:59:51.710Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.7526881720430108, "item_f1": 0.752688, "runs": 1, "total_questions": 93, "total_correct": 70, "cost_per_task_usd": 0.16370044086021504, "latency_mean_ms": 48017.408602150535, "last_run_at": "2026-08-11T11:59:53.779Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.7415730337078652, "item_f1": 0.7415729999999999, "runs": 1, "total_questions": 89, "total_correct": 66, "cost_per_task_usd": 0.6666871235955056, "latency_mean_ms": 103163.67415730337, "last_run_at": "2026-08-11T11:59:51.522Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.7171717171717171, "item_f1": 0.717172, "runs": 1, "total_questions": 99, "total_correct": 71, "cost_per_task_usd": 0.37142742424242425, "latency_mean_ms": 85480.40404040404, "last_run_at": "2026-08-11T11:58:11.074Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.7142857142857143, "item_f1": 0.714286, "runs": 1, "total_questions": 98, "total_correct": 70, "cost_per_task_usd": 0.40344476530612244, "latency_mean_ms": 76522.08163265306, "last_run_at": "2026-08-11T11:58:48.742Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.711340206185567, "item_f1": 0.71134, "runs": 1, "total_questions": 97, "total_correct": 69, "cost_per_task_usd": 0.40030438144329894, "latency_mean_ms": 120305.72164948453, "last_run_at": "2026-08-11T11:59:51.798Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.71, "item_f1": 0.71, "runs": 1, "total_questions": 100, "total_correct": 71, "cost_per_task_usd": 0.32602296000000003, "latency_mean_ms": 83956.72, "last_run_at": "2026-08-11T11:59:53.865Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.7, "item_f1": 0.7, "runs": 1, "total_questions": 100, "total_correct": 70, "cost_per_task_usd": 0.11213487000000001, "latency_mean_ms": 44045.37, "last_run_at": "2026-08-11T12:00:01.962Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.69, "item_f1": 0.69, "runs": 1, "total_questions": 100, "total_correct": 69, "cost_per_task_usd": 0.25277484, "latency_mean_ms": 67829.85, "last_run_at": "2026-08-11T11:59:51.858Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6857142857142857, "item_f1": 0.6857144705882353, "runs": 2, "total_questions": 595, "total_correct": 408, "cost_per_task_usd": 0.5376067210084033, "latency_mean_ms": 123681.8268907563, "last_run_at": "2026-08-17T06:04:35.667Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.6853932584269663, "item_f1": 0.685393, "runs": 1, "total_questions": 89, "total_correct": 61, "cost_per_task_usd": 0.1651121011235955, "latency_mean_ms": 48101.04494382023, "last_run_at": "2026-08-11T11:58:48.652Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6842105263157895, "item_f1": 0.684211, "runs": 1, "total_questions": 95, "total_correct": 65, "cost_per_task_usd": 0.38464224210526315, "latency_mean_ms": 102418.86315789474, "last_run_at": "2026-08-11T11:59:50.673Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.6635338345864662, "item_f1": 0.6635340150375939, "runs": 2, "total_questions": 532, "total_correct": 353, "cost_per_task_usd": 0.46581978383458644, "latency_mean_ms": 115865.9887218045, "last_run_at": "2026-08-17T07:48:29.587Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.66, "item_f1": 0.66, "runs": 1, "total_questions": 100, "total_correct": 66, "cost_per_task_usd": 0.15599296, "latency_mean_ms": 79347.91, "last_run_at": "2026-08-11T12:00:01.871Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.66, "item_f1": 0.66, "runs": 1, "total_questions": 100, "total_correct": 66, "cost_per_task_usd": 0.2750656, "latency_mean_ms": 77411.35, "last_run_at": "2026-08-11T11:58:52.318Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6503267973856209, "item_f1": 0.6503270326797386, "runs": 2, "total_questions": 612, "total_correct": 398, "cost_per_task_usd": 0.39914809640522875, "latency_mean_ms": 97439.42810457517, "last_run_at": "2026-08-17T05:53:21.452Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.65, "item_f1": 0.65, "runs": 1, "total_questions": 100, "total_correct": 65, "cost_per_task_usd": 0.28106902, "latency_mean_ms": 72163.67, "last_run_at": "2026-08-11T12:00:20.014Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.6428571428571429, "item_f1": 0.6428567819548872, "runs": 2, "total_questions": 532, "total_correct": 342, "cost_per_task_usd": 0.3260296954887218, "latency_mean_ms": 83209.16541353383, "last_run_at": "2026-08-17T07:48:29.612Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.64, "item_f1": 0.64, "runs": 1, "total_questions": 100, "total_correct": 64, "cost_per_task_usd": 0.1233334, "latency_mean_ms": 50328.14, "last_run_at": "2026-08-11T11:56:20.154Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.6009615384615384, "item_f1": 0.6009614567307693, "runs": 2, "total_questions": 624, "total_correct": 375, "cost_per_task_usd": 0.20370741185897437, "latency_mean_ms": 66844.17948717948, "last_run_at": "2026-08-17T06:04:26.898Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.5789473684210527, "item_f1": 0.5789475187969926, "runs": 2, "total_questions": 532, "total_correct": 308, "cost_per_task_usd": 0.15172465413533837, "latency_mean_ms": 68841.86654135339, "last_run_at": "2026-08-17T07:48:29.614Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openrouter/auto", "model_name": "Auto Router", "search_engine": "auto", "surface": "server-tool", "max_agent_turns": null, "reasoning_effort": "high", "score": 0.5245026686074721, "item_f1": 0.5245029999999999, "runs": 1, "total_questions": 2061, "total_correct": 1081, "cost_per_task_usd": 0.08189959243085881, "latency_mean_ms": 123245.16836487142, "last_run_at": "2026-09-15T18:51:27.288Z" }, { "benchmark": "search_hle", "benchmark_page_url": "https://openrouter.ai/benchmarks/hle", "model_permaslug": "openrouter/auto-beta", "model_name": "Auto Router (Beta)", "search_engine": "auto", "surface": "server-tool", "max_agent_turns": null, "reasoning_effort": "high", "score": 0.4458568738229755, "item_f1": 0.445857, "runs": 1, "total_questions": 2124, "total_correct": 947, "cost_per_task_usd": 0.0861409016007533, "latency_mean_ms": 148003.12853107345, "last_run_at": "2026-09-15T18:51:02.455Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.22, "item_f1": 0.834223, "runs": 1, "total_questions": 100, "total_correct": 22, "cost_per_task_usd": 0.16628389, "latency_mean_ms": 164892, "last_run_at": "2026-08-11T05:37:04.176Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.21, "item_f1": 0.8404900000000001, "runs": 1, "total_questions": 100, "total_correct": 21, "cost_per_task_usd": 0.83365428, "latency_mean_ms": 135725.25, "last_run_at": "2026-08-11T00:50:53.051Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.20202020202020202, "item_f1": 0.68224, "runs": 1, "total_questions": 99, "total_correct": 20, "cost_per_task_usd": 2.141136484848485, "latency_mean_ms": 150384.55555555556, "last_run_at": "2026-08-11T00:57:02.576Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.2, "item_f1": 0.771289, "runs": 1, "total_questions": 100, "total_correct": 20, "cost_per_task_usd": 0.10469292, "latency_mean_ms": 118587.16, "last_run_at": "2026-08-11T10:51:14.371Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.18, "item_f1": 0.799834, "runs": 1, "total_questions": 100, "total_correct": 18, "cost_per_task_usd": 0.06317221, "latency_mean_ms": 146626.7, "last_run_at": "2026-08-11T05:25:51.761Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.18, "item_f1": 0.814303, "runs": 1, "total_questions": 100, "total_correct": 18, "cost_per_task_usd": 0.45344653, "latency_mean_ms": 130648.49, "last_run_at": "2026-08-11T00:50:53.050Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.17346938775510204, "item_f1": 0.7213040000000001, "runs": 1, "total_questions": 98, "total_correct": 17, "cost_per_task_usd": 0.05012339795918367, "latency_mean_ms": 155838.0306122449, "last_run_at": "2026-08-11T05:37:32.838Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.17, "item_f1": 0.78417, "runs": 1, "total_questions": 100, "total_correct": 17, "cost_per_task_usd": 0.53445107, "latency_mean_ms": 164499.78, "last_run_at": "2026-08-11T11:09:29.347Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.165, "item_f1": 0.806244, "runs": 2, "total_questions": 200, "total_correct": 33, "cost_per_task_usd": 1.463678175, "latency_mean_ms": 161189.6, "last_run_at": "2026-08-17T07:48:29.610Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.16, "item_f1": 0.7799809999999999, "runs": 1, "total_questions": 100, "total_correct": 16, "cost_per_task_usd": 0.21862183999999998, "latency_mean_ms": 171969.4, "last_run_at": "2026-08-11T05:35:49.334Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.16, "item_f1": 0.81759, "runs": 1, "total_questions": 100, "total_correct": 16, "cost_per_task_usd": 0.97505473, "latency_mean_ms": 231028.03, "last_run_at": "2026-08-11T11:09:29.471Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.15, "item_f1": 0.7447750000000001, "runs": 1, "total_questions": 100, "total_correct": 15, "cost_per_task_usd": 0.14567118, "latency_mean_ms": 138409.41, "last_run_at": "2026-08-11T10:55:06.366Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.135678391959799, "item_f1": 0.6593649798994975, "runs": 2, "total_questions": 199, "total_correct": 27, "cost_per_task_usd": 4.423114663316583, "latency_mean_ms": 209530.93467336684, "last_run_at": "2026-08-17T05:48:07.078Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.135, "item_f1": 0.7776430000000001, "runs": 2, "total_questions": 200, "total_correct": 27, "cost_per_task_usd": 0.59830761, "latency_mean_ms": 154010.185, "last_run_at": "2026-08-17T07:48:29.614Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.13131313131313133, "item_f1": 0.677881, "runs": 1, "total_questions": 99, "total_correct": 13, "cost_per_task_usd": 0.7363743535353535, "latency_mean_ms": 110382.85858585859, "last_run_at": "2026-08-11T00:55:16.334Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "native", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.13, "item_f1": 0.768655, "runs": 1, "total_questions": 100, "total_correct": 13, "cost_per_task_usd": 0.34829433000000004, "latency_mean_ms": 161520.1, "last_run_at": "2026-08-11T11:09:29.183Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.13, "item_f1": 0.774054, "runs": 2, "total_questions": 200, "total_correct": 26, "cost_per_task_usd": 0.32877032, "latency_mean_ms": 122325.835, "last_run_at": "2026-08-17T07:48:29.620Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.13, "item_f1": 0.797676, "runs": 1, "total_questions": 100, "total_correct": 13, "cost_per_task_usd": 0.26298521999999996, "latency_mean_ms": 114810.94, "last_run_at": "2026-08-11T00:50:52.014Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.12121212121212122, "item_f1": 0.657525, "runs": 1, "total_questions": 99, "total_correct": 12, "cost_per_task_usd": 0.6157617373737374, "latency_mean_ms": 122930.72727272728, "last_run_at": "2026-08-10T23:33:11.434Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.12121212121212122, "item_f1": 0.669667, "runs": 1, "total_questions": 99, "total_correct": 12, "cost_per_task_usd": 1.8469389292929295, "latency_mean_ms": 165973.84848484848, "last_run_at": "2026-08-10T23:35:57.895Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.12, "item_f1": 0.756606, "runs": 1, "total_questions": 100, "total_correct": 12, "cost_per_task_usd": 0.07468046, "latency_mean_ms": 146234.85, "last_run_at": "2026-08-11T05:15:24.154Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.11557788944723618, "item_f1": 0.743618, "runs": 1, "total_questions": 199, "total_correct": 23, "cost_per_task_usd": 0.4659304522613066, "latency_mean_ms": 187392.99497487437, "last_run_at": "2026-08-10T22:20:35.830Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.11557788944723618, "item_f1": 0.7735449999999999, "runs": 1, "total_questions": 199, "total_correct": 23, "cost_per_task_usd": 0.8579122814070352, "latency_mean_ms": 204930.56281407035, "last_run_at": "2026-08-10T22:20:35.647Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.115, "item_f1": 0.6517149999999999, "runs": 2, "total_questions": 200, "total_correct": 23, "cost_per_task_usd": 0.95424031, "latency_mean_ms": 124147.08, "last_run_at": "2026-08-17T05:48:07.453Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "xhigh", "score": 0.115, "item_f1": 0.7439695000000001, "runs": 2, "total_questions": 200, "total_correct": 23, "cost_per_task_usd": 0.120836625, "latency_mean_ms": 181141.955, "last_run_at": "2026-08-18T02:37:59.398Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "xhigh", "score": 0.11, "item_f1": 0.728517, "runs": 2, "total_questions": 200, "total_correct": 22, "cost_per_task_usd": 0.05848085, "latency_mean_ms": 149190.745, "last_run_at": "2026-08-18T02:36:37.639Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.11, "item_f1": 0.710925, "runs": 1, "total_questions": 100, "total_correct": 11, "cost_per_task_usd": 0.0365038, "latency_mean_ms": 144244.39, "last_run_at": "2026-08-11T04:57:36.341Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "perplexity", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.1, "item_f1": 0.6582049999999999, "runs": 1, "total_questions": 100, "total_correct": 10, "cost_per_task_usd": 0.23463712, "latency_mean_ms": 70005.74, "last_run_at": "2026-08-11T00:57:10.783Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 25, "reasoning_effort": "high", "score": 0.0979381443298969, "item_f1": 0.7123584742268041, "runs": 2, "total_questions": 194, "total_correct": 19, "cost_per_task_usd": 0.08851609278350517, "latency_mean_ms": 195671.3556701031, "last_run_at": "2026-08-18T02:36:38.739Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.09183673469387756, "item_f1": 0.6628930000000001, "runs": 1, "total_questions": 98, "total_correct": 9, "cost_per_task_usd": 0.05909258163265306, "latency_mean_ms": 168319.4081632653, "last_run_at": "2026-08-11T05:36:02.791Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-sol", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.08542713567839195, "item_f1": 0.681502, "runs": 1, "total_questions": 199, "total_correct": 17, "cost_per_task_usd": 0.28785209045226134, "latency_mean_ms": 167107.37688442212, "last_run_at": "2026-08-10T20:39:10.725Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.085, "item_f1": 0.679267, "runs": 2, "total_questions": 200, "total_correct": 17, "cost_per_task_usd": 0.039141455, "latency_mean_ms": 152748.925, "last_run_at": "2026-08-18T02:36:38.654Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.08, "item_f1": 0.6167435, "runs": 2, "total_questions": 200, "total_correct": 16, "cost_per_task_usd": 0.26071565, "latency_mean_ms": 77328.94, "last_run_at": "2026-08-17T05:48:07.356Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 5, "reasoning_effort": "high", "score": 0.08, "item_f1": 0.6616725, "runs": 2, "total_questions": 200, "total_correct": 16, "cost_per_task_usd": 0.04025679, "latency_mean_ms": 171011.02, "last_run_at": "2026-08-18T02:39:27.233Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "openai/gpt-5.6-luna", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "xhigh", "score": 0.08, "item_f1": 0.691202, "runs": 1, "total_questions": 100, "total_correct": 8, "cost_per_task_usd": 0.03828898, "latency_mean_ms": 145927.21, "last_run_at": "2026-08-11T04:53:53.429Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "anthropic/claude-opus-5", "model_name": null, "search_engine": "exa", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.05527638190954774, "item_f1": 0.596167, "runs": 1, "total_questions": 199, "total_correct": 11, "cost_per_task_usd": 0.2636326984924623, "latency_mean_ms": 89750.92462311558, "last_run_at": "2026-08-10T20:39:10.641Z" }, { "benchmark": "search_widesearch", "benchmark_page_url": "https://openrouter.ai/benchmarks/widesearch", "model_permaslug": "deepseek/deepseek-v4-flash-0731", "model_name": null, "search_engine": "parallel", "surface": "server-tool", "max_agent_turns": 1, "reasoning_effort": "high", "score": 0.04, "item_f1": 0.5940679999999999, "runs": 1, "total_questions": 100, "total_correct": 4, "cost_per_task_usd": 0.01869195, "latency_mean_ms": 110308.85, "last_run_at": "2026-08-18T02:36:37.725Z" } ] }