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
51GPT-5.6 SolOpenRouter harness, openai/gpt-5.6-sol-pro-20260709, auto-routing, public 90-day aggregate, pro· OpenRouter benchmark exports94.1%
52Gemini 3.1 Pro Preview· Epoch AI Benchmarking Hub data94.1%
53GPT-5.6 SolDeepSeek V4.1 launch chart · exact public value · setting not further specified· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.94.1%
54GPT-5.6 SolPass@1 / max effort· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.94.1%
55GPT-5.6 SolQwen3.8 官方模型卡对照表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.94.1%
56Grok 4.6high· Epoch AI Benchmarking Hub data94%
57GPT 5.5OpenRouter harness, openai/gpt-5.5-20260423, auto-routing, public 90-day aggregate· OpenRouter benchmark exports93.9%
58GPT-5.6 SolOpenRouter harness, openai/gpt-5.6-sol-pro-20260709, Azure, public 90-day aggregate, pro· OpenRouter benchmark exports93.9%
59GPT-5.6 SolOpenRouter harness, openai/gpt-5.6-sol-pro-20260709, OpenAI, public 90-day aggregate, pro· OpenRouter benchmark exports93.9%
60Claude Fable 5.2max· Epoch AI Benchmarking Hub data93.9%
61Claude Opus 5max· Epoch AI Benchmarking Hub data93.9%
62Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Sail Research, public 90-day aggregate· OpenRouter benchmark exports93.9%
63Gemini 3.7 FlashOpenRouter harness, google/gemini-3.7-flash-20260813, auto-routing, public 90-day aggregate· OpenRouter benchmark exports93.9%
64GPT-6.1 SolOpenRouter harness, openai/gpt-6.1-sol-20260929, Azure, public 90-day aggregate· OpenRouter benchmark exports93.8%
65Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Nebius, public 90-day aggregate· OpenRouter benchmark exports93.8%
66Claude Fable 5.1OpenAI launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.7%
67Claude Opus 5OpenAI launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.7%
68Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, River, public 90-day aggregate· OpenRouter benchmark exports93.7%
69GPT 5.5OpenRouter harness, openai/gpt-5.5-20260423, Azure, public 90-day aggregate· OpenRouter benchmark exports93.7%
70Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, auto-routing, public 90-day aggregate· OpenRouter benchmark exports93.6%
71GPT 5.5OpenRouter harness, openai/gpt-5.5-20260423, OpenAI, public 90-day aggregate· OpenRouter benchmark exports93.6%
72Claude Opus 4.8Opus 4.8 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.6%
73GPT 5.5OpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.6%
74GPT 5.5xhigh· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.6%
75Kimi K3max effort · temp 1.0 · top_p .95· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.5%
76GPT-5.6 Solmax· Epoch AI Benchmarking Hub data93.5%
77Grok 4.5high· Epoch AI Benchmarking Hub data93.4%
78Claude Opus 5DeepSeek V4.1 launch chart · exact public value · setting not further specified· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.4%
79Claude Opus 5Pass@1 / max effort· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.4%
80Gemini 3.6 FlashOpenRouter harness, google/gemini-3.6-flash-20260721, Google, public 90-day aggregate· OpenRouter benchmark exports93.4%
81GPT-5.6 Terramax· Epoch AI Benchmarking Hub data93.3%
82GPT 5.4xhigh· Epoch AI Benchmarking Hub data93.3%
83Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Decart, public 90-day aggregate· OpenRouter benchmark exports93.3%
84Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, DigitalOcean, public 90-day aggregate· OpenRouter benchmark exports93.3%
85GPT 5.2Opus 4.6 System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.2%
86GPT-5.2 Prono tools · xhigh· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.2%
87GPT-5.2 ProOpenAI GPT-5.4 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.2%
88DeepSeek V4.1 Flash 1017max· Epoch AI Benchmarking Hub data93.2%
89Grok 4.6xhigh· Epoch AI Benchmarking Hub data93.2%
90Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Modal, public 90-day aggregate· OpenRouter benchmark exports93.2%
91Doubao Seed Evolvinghigh· Epoch AI Benchmarking Hub data93.1%
92Kimi K3max· Epoch AI Benchmarking Hub data93.1%
93Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Parasail, public 90-day aggregate· OpenRouter benchmark exports93.1%
94Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, InferenceNet, public 90-day aggregate· OpenRouter benchmark exports93.1%
95Kimi K3OpenRouter harness, moonshotai/kimi-k3-20260715, Io Net, public 90-day aggregate· OpenRouter benchmark exports93.1%
96Claude Sonnet 5.5OpenRouter harness, anthropic/claude-sonnet-5.5-20260928, Google, public 90-day aggregate· OpenRouter benchmark exports93%
97Gemini 3.6 FlashOpenRouter harness, google/gemini-3.6-flash-20260721, Google AI Studio, public 90-day aggregate· OpenRouter benchmark exports93%
98Claude Sonnet 5.5OpenRouter harness, anthropic/claude-sonnet-5.5-20260928, Claude Platform on AWS, public 90-day aggregate· OpenRouter benchmark exports93%
99Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Moonshot AI, public 90-day aggregate· OpenRouter benchmark exports92.9%
100Kimi K3.1 ProOpenRouter harness, moonshotai/kimi-k3.1-pro-20261023, Morph, public 90-day aggregate· OpenRouter benchmark exports92.9%

Showing 51–100 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)