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 | |
|---|---|---|
| 1451 | Qwen3 14BOpenRouter harness, qwen/qwen3-14b-04-28, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 59.2% |
| 1452 | GLM 4.7 FlashOpenRouter harness, z-ai/glm-4.7-flash-20260119, Novita, public 90-day aggregate· OpenRouter benchmark exports | 59.1% |
| 1453 | GPT OSS 120BOpenRouter harness, openai/gpt-oss-120b, Cerebras, public 90-day aggregate· OpenRouter benchmark exports | 58.8% |
| 1454 | GPT OSS 20bOpenRouter harness, openai/gpt-oss-20b, Novita, public 90-day aggregate· OpenRouter benchmark exports | 58.5% |
| 1455 | GPT OSS 20bofficial model card Table 3 · reasoning low · with tools; Table 3 does not name the exact tool · pass@1· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 58% |
| 1456 | GPT OSS 20bOpenRouter harness, openai/gpt-oss-20b, Google, public 90-day aggregate· OpenRouter benchmark exports | 57.7% |
| 1457 | GPT-5 nanolow· Epoch AI Benchmarking Hub data | 57.6% |
| 1458 | gemini-1.5-pro-002· Epoch AI Benchmarking Hub data | 57.2% |
| 1459 | Llama 4 Scoutinstruction-tuned table· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.2% |
| 1460 | Gemini 2.0 Flash Experimentalthinking· Epoch AI Benchmarking Hub data | 57.1% |
| 1461 | GPT OSS 20b· OpenEvals Leaderboard data· Reported by the model's developer. The source relaying it did not verify it. | 56.8% |
| 1462 | GPT OSS 20bofficial model card Table 3 · reasoning low · no tools · pass@1· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.8% |
| 1463 | Qwen3 8B· Epoch AI Benchmarking Hub data | 56.8% |
| 1464 | Qwen Max· Epoch AI Benchmarking Hub data | 56.1% |
| 1465 | Gemini 2.5 Flash-LiteOpenRouter harness, google/gemini-2.5-flash-lite, Google, public 90-day aggregate· OpenRouter benchmark exports | 56.1% |
| 1466 | Magistral Small 25092506· Epoch AI Benchmarking Hub data | 56.1% |
| 1467 | Phi 4· Epoch AI Benchmarking Hub data | 56.1% |
| 1468 | DeepSeek R1 Distill Llama 70b· Epoch AI Benchmarking Hub data | 55.7% |
| 1469 | Qwen3 32BOpenRouter harness, qwen/qwen3-32b-04-28, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 55.7% |
| 1470 | Qwen3 30B A3B Instruct· Epoch AI Benchmarking Hub data | 55.6% |
| 1471 | GPT 5.4 Nanonone· Epoch AI Benchmarking Hub data | 55.6% |
| 1472 | Claude 3.5 Sonnet· Epoch AI Benchmarking Hub data | 55.3% |
| 1473 | Gemini 2.5 Flash-LiteOpenRouter harness, google/gemini-2.5-flash-lite, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 55.2% |
| 1474 | GLM 5OpenRouter harness, z-ai/glm-5-20260211, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 54.9% |
| 1475 | Qwen3 32BOpenRouter harness, qwen/qwen3-32b-04-28, Tenstorrent, public 90-day aggregate· OpenRouter benchmark exports | 54.7% |
| 1476 | DeepSeek V4 ProOpenRouter harness, deepseek/deepseek-v4-pro-20260423, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 54.3% |
| 1477 | Qwen3 32Bnone· Epoch AI Benchmarking Hub data | 54.1% |
| 1478 | Qwen3 VL 8B InstructOpenRouter harness, qwen/qwen3-vl-8b-instruct, PrimeIntellect, public 90-day aggregate· OpenRouter benchmark exports | 53.9% |
| 1479 | Grok 2· Epoch AI Benchmarking Hub data | 53.8% |
| 1480 | DeepSeek ChatOpenRouter harness, deepseek/deepseek-chat-v3, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 53.7% |
| 1481 | Qwen3 14Bnone· Epoch AI Benchmarking Hub data | 53.4% |
| 1482 | GPT OSS 20blow· Epoch AI Benchmarking Hub data | 53.2% |
| 1483 | GPT 4oOpenRouter harness, openai/gpt-4o-2024-08-06, OpenAI, public 90-day aggregate· OpenRouter benchmark exports | 52.8% |
| 1484 | GPT 4oOpenRouter harness, openai/gpt-4o, OpenAI, public 90-day aggregate· OpenRouter benchmark exports | 52.7% |
| 1485 | Qwen3 4B· Epoch AI Benchmarking Hub data | 52.3% |
| 1486 | DeepSeek V3OpenRouter harness, deepseek/deepseek-chat-v3-0324, GMICloud, public 90-day aggregate· OpenRouter benchmark exports | 52.1% |
| 1487 | GPT 4oOpenRouter harness, openai/gpt-4o-2024-08-06, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 52% |
| 1488 | Llama 4 Scout· Epoch AI Benchmarking Hub data | 51.8% |
| 1489 | GPT 4oOpenRouter harness, openai/gpt-4o, Azure, public 90-day aggregate· OpenRouter benchmark exports | 51.7% |
| 1490 | GPT 4oOpenRouter harness, openai/gpt-4o-2024-08-06, Azure, public 90-day aggregate· OpenRouter benchmark exports | 51.5% |
| 1491 | Mistral Large2411· Epoch AI Benchmarking Hub data | 51.3% |
| 1492 | DeepSeek V3OpenRouter harness, deepseek/deepseek-chat-v3-0324, DeepInfra, public 90-day aggregate· OpenRouter benchmark exports | 51.3% |
| 1493 | GLM 4.7 FlashOpenRouter harness, z-ai/glm-4.7-flash-20260119, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 51.1% |
| 1494 | GPT 4oOpenRouter harness, openai/gpt-4o, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 51.1% |
| 1495 | Llama 3.1 405B Instruct· Epoch AI Benchmarking Hub data | 50.9% |
| 1496 | GPT-4.1 nanoOpenRouter harness, openai/gpt-4.1-nano-2025-04-14, OpenAI, public 90-day aggregate· OpenRouter benchmark exports | 50.5% |
| 1497 | Ling 3.0 Flash FinOpenRouter harness, inclusionai/ling-3.0-flash-fin-20260827, auto-routing, public 90-day aggregate· OpenRouter benchmark exports | 50.5% |
| 1498 | Qwen3 30B A3Bnone· Epoch AI Benchmarking Hub data | 50.4% |
| 1499 | O1 Preview· Epoch AI Benchmarking Hub data | 50.3% |
| 1500 | GPT-4.1 nanoOpenRouter harness, openai/gpt-4.1-nano-2025-04-14, Azure, public 90-day aggregate· OpenRouter benchmark exports | 50% |
Showing 1,451–1,500 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)