Skip to content

Benchmark

GPQA Diamond

Science benchmarks measure how well models answer science questions. Scores are percentages. Higher is better.

Models and systems measured
282
Scores recorded
1623
Sources
4
IQ weight
1.72

Score distribution

79.3%
Mean
84.5%
Median
15.7%
Std dev
56.8%
10th percentile
92.2%
90th percentile
0.168%
Min
0.168%96.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.72 for chat models, measured over 272 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or system
851Qwen3.6 35B A3BOpenRouter harness, qwen/qwen3.6-35b-a3b-20260415, SiliconFlow, public 90-day aggregate· OpenRouter benchmark exports84.1%
852Gemini 3.5 Flash LiteOpenRouter harness, google/gemini-3.5-flash-lite-20260721, Google AI Studio, public 90-day aggregate· OpenRouter benchmark exports84.1%
853GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Together, public 90-day aggregate· OpenRouter benchmark exports84.1%
854GLM 5.3 FlashOpenRouter harness, z-ai/glm-5.3-flash-20260826, auto-routing, public 90-day aggregate· OpenRouter benchmark exports84.1%
855GLM 5.3 FlashOpenRouter harness, z-ai/glm-5.3-flash-20260826, Fireworks, public 90-day aggregate· OpenRouter benchmark exports84.1%
856GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Sail Research, public 90-day aggregate· OpenRouter benchmark exports84%
857GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Reka, public 90-day aggregate· OpenRouter benchmark exports84%
858GLM 5.1OpenRouter harness, z-ai/glm-5.1-20260406, Chutes, public 90-day aggregate· OpenRouter benchmark exports84%
859Qwen3.5 122B-A10BOpenRouter harness, qwen/qwen3.5-122b-a10b-20260224, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports84%
860GLM 5.2OpenRouter harness, z-ai/glm-5.2-20260616, Mistral, public 90-day aggregate· OpenRouter benchmark exports84%
861Gemini 3.5 Flash LiteOpenRouter harness, google/gemini-3.5-flash-lite-20260721, Google, public 90-day aggregate· OpenRouter benchmark exports84%
862DeepSeek V4.1 FlashOpenRouter harness, deepseek/deepseek-v4.1-flash-20260910, Io Net, public 90-day aggregate· OpenRouter benchmark exports84%
863GLM 4.6OpenRouter harness, z-ai/glm-4.6, Novita, public 90-day aggregate· OpenRouter benchmark exports84%
864DeepSeek V4.1 Flash 1017OpenRouter harness, deepseek/deepseek-v4.1-flash-20261017, Reka, public 90-day aggregate· OpenRouter benchmark exports83.9%
865MiMo V2.5 ProOpenRouter harness, xiaomi/mimo-v2.5-pro-20260422, DigitalOcean, public 90-day aggregate· OpenRouter benchmark exports83.9%
866GLM 5.2OpenRouter harness, z-ai/glm-5.2-20260616, CoreWeave, public 90-day aggregate· OpenRouter benchmark exports83.9%
867GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Nebius, public 90-day aggregate· OpenRouter benchmark exports83.9%
868GLM 4.7OpenRouter harness, z-ai/glm-4.7-20251222, Novita, public 90-day aggregate· OpenRouter benchmark exports83.9%
869Qwen3.5 122B-A10BOpenRouter harness, qwen/qwen3.5-122b-a10b-20260224, AtlasCloud, public 90-day aggregate· OpenRouter benchmark exports83.9%
870Kimi K2.7 CodeOpenRouter harness, moonshotai/kimi-k2.7-code-20260612, StreamLake, public 90-day aggregate· OpenRouter benchmark exports83.9%
871Qwen 3.8 27BOpenRouter harness, qwen/qwen3.8-27b-20260814, StreamLake, public 90-day aggregate· OpenRouter benchmark exports83.9%
872GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Novita, public 90-day aggregate· OpenRouter benchmark exports83.9%
873Gemini 2.5 Pro· Epoch AI Benchmarking Hub data83.8%
874Qwen3.6 35B A3B· Epoch AI Benchmarking Hub data83.8%
875GPT-5.4 miniOpenRouter harness, openai/gpt-5.4-mini-20260317, Azure, public 90-day aggregate· OpenRouter benchmark exports83.8%
876Claude Sonnet 4extended thinking · parallel test-time compute· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.83.8%
877GLM 5.2OpenRouter harness, z-ai/glm-5.2-20260616, Parasail, public 90-day aggregate· OpenRouter benchmark exports83.8%
878GLM 5.2OpenRouter harness, z-ai/glm-5.2-20260616, InferenceNet, public 90-day aggregate· OpenRouter benchmark exports83.8%
879GLM 5.2OpenRouter harness, z-ai/glm-5.2-20260616, Morph, public 90-day aggregate· OpenRouter benchmark exports83.8%
880GLM 4.7OpenRouter harness, z-ai/glm-4.7-20251222, Google, public 90-day aggregate· OpenRouter benchmark exports83.8%
881GPT-5.4 miniOpenRouter harness, openai/gpt-5.4-mini-20260317, auto-routing, public 90-day aggregate· OpenRouter benchmark exports83.8%
882Qwen3.6 35B A3BOpenRouter harness, qwen/qwen3.6-35b-a3b-20260415, AkashML, public 90-day aggregate· OpenRouter benchmark exports83.7%
883GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, Modal, public 90-day aggregate· OpenRouter benchmark exports83.7%
884MiMo V2 Flashofficial release/model-card table· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.83.7%
885DeepSeek V4 FlashOpenRouter harness, deepseek/deepseek-v4-flash-20260423, StreamLake, public 90-day aggregate· OpenRouter benchmark exports83.7%
886GLM 4.7OpenRouter harness, z-ai/glm-4.7-20251222, AtlasCloud, public 90-day aggregate· OpenRouter benchmark exports83.7%
887Qwen3.6 35B A3BOpenRouter harness, qwen/qwen3.6-35b-a3b-20260415, Reka, public 90-day aggregate· OpenRouter benchmark exports83.6%
888GLM 5.3 FlashOpenRouter harness, z-ai/glm-5.3-flash-20260826, Reka, public 90-day aggregate· OpenRouter benchmark exports83.6%
889GPT-5.4 minihigh· Epoch AI Benchmarking Hub data83.6%
890Qwen3.6 27BOpenRouter harness, qwen/qwen3.6-27b-20260422, SiliconFlow, public 90-day aggregate· OpenRouter benchmark exports83.6%
891Ling 3.0 Flash VLOpenRouter harness, inclusionai/ling-3.0-flash-vl-20260910, Novita, public 90-day aggregate· OpenRouter benchmark exports83.6%
892GLM 5.1OpenRouter harness, z-ai/glm-5.1-20260406, Z.AI, public 90-day aggregate· OpenRouter benchmark exports83.6%
893Ling 3.0 Flash VLOpenRouter harness, inclusionai/ling-3.0-flash-vl-20260910, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports83.5%
894Gemma 4 31B ITOpenRouter harness, google/gemma-4-31b-it-20260402, Phala, public 90-day aggregate· OpenRouter benchmark exports83.5%
895Muse Glimmer 30BAA · high reasoning· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.83.5%
896Step 3.5 Flash· OpenEvals Leaderboard data· Reported by the model's developer. The source relaying it did not verify it.83.5%
897Step 3.5 FlashStep 3.5 technical report · Table 5 · vanilla· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.83.5%
898GLM 5.3OpenRouter harness, z-ai/glm-5.3-20260816, AkashML, public 90-day aggregate· OpenRouter benchmark exports83.5%
899MiniMax M3OpenRouter harness, minimax/minimax-m3-20260531, Mara, public 90-day aggregate· OpenRouter benchmark exports83.5%
900Qwen 3.8 27BOpenRouter harness, qwen/qwen3.8-27b-20260814, Modal, public 90-day aggregate· OpenRouter benchmark exports83.4%

Showing 851–900 of 1,621 results.

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or system
MiMo V2 Flash Baseofficial Base table · 5-shot· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.55.1%
Qwen3.8-2.4T-A95Blifecycle-page core-benchmark claim· Reported by Alibaba Cloud Model Studio· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.92.6%

Tokens the model generated per task, as measured by the source. Comparable within this benchmark only.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence) · OpenEvals/leaderboard-data, MIT (licence)