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
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 | |
|---|---|---|
| 1601 | Llama 3.1 8B InstructOpenRouter harness, meta-llama/llama-3.1-8b-instruct, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 26.1% |
| 1602 | Llama 3 8B Instruct· Epoch AI Benchmarking Hub data | 26.1% |
| 1603 | Llama 3.1 8B InstructOpenRouter harness, meta-llama/llama-3.1-8b-instruct, Novita, public 90-day aggregate· OpenRouter benchmark exports | 25.9% |
| 1604 | Ministral 3B· Epoch AI Benchmarking Hub data | 25.3% |
| 1605 | DeepSeek LLM 67B Chat· Epoch AI Benchmarking Hub data | 24.6% |
| 1606 | Llama 3.2 1B Instruct· Epoch AI Benchmarking Hub data | 23.9% |
| 1607 | Gemma 3 4B IT· Epoch AI Benchmarking Hub data | 23.2% |
| 1608 | Gemini Nano Banana 2.1OpenRouter harness, google/gemini-nano-banana-2.1-20261006, Google, public 90-day aggregate· OpenRouter benchmark exports | 21.4% |
| 1609 | Gemma 3 1b it· Epoch AI Benchmarking Hub data | 19.9% |
| 1610 | Ling 3.0 Flash FinOpenRouter harness, inclusionai/ling-3.0-flash-fin-20260827, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 17.2% |
| 1611 | Mistral (7B) Instruct v0.3· Epoch AI Benchmarking Hub data | 15.2% |
| 1612 | Yi 34B Chat· Epoch AI Benchmarking Hub data | 14.7% |
| 1613 | open-mistral-7b· Epoch AI Benchmarking Hub data | 13.2% |
| 1614 | Qwen3.5 0.8B· OpenEvals Leaderboard data· Reported by the model's developer. The source relaying it did not verify it. | 11.9% |
| 1615 | Gemini 3.1 Flash-Lite ImageOpenRouter harness, google/gemini-3.1-flash-lite-image-20260630, Google AI Studio, public 90-day aggregate· OpenRouter benchmark exports | 11.6% |
| 1616 | Gemini 3.1 Flash-Lite ImageOpenRouter harness, google/gemini-3.1-flash-lite-image-20260630, Google, public 90-day aggregate· OpenRouter benchmark exports | 11.3% |
| 1617 | Gemini 3.1 Flash-Lite ImageOpenRouter harness, google/gemini-3.1-flash-lite-image-20260630, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 10.3% |
| 1618 | DeepSeek R1 0528 Qwen3 8B· Epoch AI Benchmarking Hub data | 9.3% |
| 1619 | DeepSeek V3OpenRouter harness, deepseek/deepseek-chat-v3-0324, Novita, public 90-day aggregate· OpenRouter benchmark exports | 8.8% |
| 1620 | DeepSeek V3.1OpenRouter harness, deepseek/deepseek-chat-v3.1, SiliconFlow, public 90-day aggregate· OpenRouter benchmark exports | 0.168% |
| 1621 | Llama 3.1 8B InstructOpenRouter harness, meta-llama/llama-3.1-8b-instruct, Cloudflare, public 90-day aggregate· OpenRouter benchmark exports | 0.168% |
Showing 1,601–1,621 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)