Knowledge
SimpleQA Leaderboard
SimpleQA measures factual accuracy on short, fact-seeking questions with a single correct answer — directly probing how often a model is right versus confidently wrong (hallucination) on simple facts.
Source: epoch15 open models ranked+63 proprietaryData through Sep 2026
All models ranked on SimpleQA
Proprietary / closed models are shown dimmed — you can't run them locally, but they show where the open field stands.
| # | Model | Score |
|---|---|---|
| 1 | GPT 6 Astra Max · proprietary | 75.6% |
| 2 | Gemini 3.1 Pro Preview (high) · proprietary | 73.5% |
| 3 | Claude Fable 5.1 Max · proprietary | 70.8% |
| 4 | Claude Fable 5 (xhigh) · proprietary | 70.7% |
| 5 | GPT 5.6 Sol Max · proprietary | 69.7% |
| 6 | Gemini 3.7 Flash (high) · proprietary | 69.2% |
| 7 | Gemini 3 Flash Preview (high) · proprietary | 66.8% |
| 8 | Gemini 3.5 Flash (high) · proprietary | 66.2% |
| 9 | Gemini 3.6 Flash (high) · proprietary | 66.2% |
| 10 | GPT 5.5 (xhigh) · proprietary | 63.0% |
| 11 | Muse Spark 1.2 (xhigh) · proprietary | 60.3% |
| 12 | Claude Opus 5 Max · proprietary | 59.9% |
| 13 | Muse Spark 1.1 · proprietary | 57.8% |
| 14 | Qwen3.7 Max · proprietary | 55.8% |
| 15 | Claude Opus 4.8 Max · proprietary | 53.0% |
| 16 | DeepSeek V4 Pro 0813 · 1650.5B | 52.9% |
| 17 | Qwen3.6 Max Preview · proprietary | 52.0% |
| 18 | Claude Opus 4.7 (xhigh) · proprietary | 51.7% |
| 19 | Kimi K3 · 2779.9B | 50.6% |
| 20 | GPT 5 (Aug 07, 2025, high) · proprietary | 50.1% |
| 21 | O3 (Apr 16, 2025, high) · proprietary | 49.4% |
| 22 | Grok 4.6 (high) · proprietary | 49.3% |
| 23 | Grok 4.6 (xhigh) · proprietary | 48.9% |
| 24 | Qwen3 Max (Sep 23, 2025) · proprietary | 48.8% |
| 25 | Grok 4.5 (high) · proprietary | 48.3% |
| 26 | GPT 5.1 (Nov 13, 2025, high) · proprietary | 48.0% |
| 27 | Claude Opus 4.6 Max · proprietary | 47.0% |
| 28 | DeepSeek V4 Pro · 1598.8B | 47.0% |
| 29 | GPT 5.4 Pro (Mar 05, 2026, xhigh) · proprietary | 46.3% |
| 30 | Qwen3.8 Max (xhigh) · proprietary | 45.8% |
| 31 | Claude Opus 4.5 (Nov 01, 2025, 32K) · proprietary | 45.7% |
| 32 | GPT 5.4 (Mar 05, 2026, xhigh) · proprietary | 45.1% |
| 33 | Qwen3.6 Plus · proprietary | 44.1% |
| 34 | GPT 5.6 Terra Max · proprietary | 43.2% |
| 35 | O1 (Dec 17, 2024, high) · proprietary | 41.1% |
| 36 | GLM 5.3 · 753.3B | 41.0% |
| 37 | GPT 5.6 Luna Max · proprietary | 41.0% |
| 38 | Qwen3 235B A22B Thinking 2507 · 235.1B | 40.4% |
| 39 | Inkling · 952.4B | 40.3% |
| 40 | GPT 5.2 (Dec 11, 2025, xhigh) · proprietary | 37.1% |
| 41 | Kimi K2.7 Code · 1026.9B | 36.5% |
| 42 | Claude Sonnet 4.6 (high) · proprietary | 35.5% |
| 43 | Kimi K2.6 · 1026.9B | 34.9% |
| 44 | GPT 5.2 (Dec 11, 2025, high) · proprietary | 34.3% |
| 45 | Kimi K2.5 · 1026.9B | 34.3% |
| 46 | GLM 5.2 · 753.3B | 34.2% |
| 47 | GLM 5.1 · 753.9B | 34.0% |
| 48 | Claude Sonnet 5 Max · proprietary | 33.7% |
| 49 | DeepSeek V4 Flash 0731 · 304.2B | 33.6% |
| 50 | Grok 4.3 (high) · proprietary | 33.2% |
| 51 | Claude Sonnet 5 (xhigh) · proprietary | 32.9% |
| 52 | Claude Sonnet 4.6 Max · proprietary | 32.8% |
| 53 | GPT 5.2 (Dec 11, 2025, low) · proprietary | 32.8% |
| 54 | GPT 5.2 (Dec 11, 2025, medium) · proprietary | 32.7% |
| 55 | GLM 4.7 · 358.3B | 32.2% |
| 56 | GPT 4.1 (Apr 14, 2025) · proprietary | 31.1% |
| 57 | Claude Sonnet 4.5 (Sep 29, 2025, 59K) · proprietary | 30.7% |
| 58 | Grok 4.20 0309 Reasoning · proprietary | 30.2% |
| 59 | GPT 5.4 Mini (Mar 17, 2026, high) · proprietary | 29.4% |
| 60 | GPT 4o (Aug 06, 2024) · proprietary | 26.0% |
| 61 | Qwen3.5 Plus · proprietary | 25.4% |
| 62 | Claude Sonnet 4.5 (Sep 29, 2025) · proprietary | 23.7% |
| 63 | GPT 5 Mini (Aug 07, 2025, high) · proprietary | 21.6% |
| 64 | Qwen3.5 Flash · proprietary | 20.3% |
| 65 | O4 Mini (Apr 16, 2025, low) · proprietary | 19.6% |
| 66 | Inkling Small · 266.0B | 19.1% |
| 67 | O4 Mini (Apr 16, 2025, high) · proprietary | 18.8% |
| 68 | Qwen3.6 Flash · proprietary | 15.9% |
| 69 | O3 Mini (Jan 31, 2025, high) · proprietary | 15.3% |
| 70 | Claude Haiku 4.5 (Oct 01, 2025) · proprietary | 13.2% |
| 71 | GPT 4.1 Mini (Apr 14, 2025) · proprietary | 12.7% |
| 72 | Claude 3 Opus (Feb 29, 2024) · proprietary | 12.6% |
| 73 | Claude Haiku 4.5 (Oct 01, 2025, 32K) · proprietary | 12.6% |
| 74 | GPT 5 Nano (Aug 07, 2025, high) · proprietary | 11.7% |
| 75 | GPT 5.4 Nano (Mar 17, 2026, high) · proprietary | 11.7% |
| 76 | Gemma 4 31B IT · 31.3B | 10.4% |
| 77 | GPT 4o Mini (Jul 18, 2024) · proprietary | 8.3% |
| 78 | GPT 4.1 Nano (Apr 14, 2025) · proprietary | 6.0% |
Score vs model size
Which models give the most quality for their size — the ones worth running locally.
- Gemma 4 31B IT, 31B, score 10.4% — on the efficiency frontier (best score at its size or smaller).
- Qwen3 235B A22B Thinking 2507, 235B, score 40.4% — on the efficiency frontier (best score at its size or smaller).
- GLM 5.3, 753B, score 41.0% — on the efficiency frontier (best score at its size or smaller).
- DeepSeek V4 Pro, 1.6T, score 47.0% — on the efficiency frontier (best score at its size or smaller).
- DeepSeek V4 Pro 0813, 1.7T, score 52.9% — on the efficiency frontier (best score at its size or smaller).
SimpleQA: frequently asked questions
- What is the best open LLM on SimpleQA?
- DeepSeek V4 Pro 0813 is the top open model on SimpleQA, scoring 52.9%. Among all models tested — including proprietary ones — it ranks #16. The top model overall is GPT 6 Astra Max (OpenAI) at 75.6%.
- What's the best SimpleQA model you can run on a 24 GB GPU?
- Gemma 4 31B IT is the highest-scoring open model that fits in 24 GB at 4-bit quantization (about 17 GB), scoring 10.4% on SimpleQA.
- Can open models match proprietary models on SimpleQA?
- Not quite on SimpleQA: the strongest proprietary model (GPT 6 Astra Max) scores 75.6%, ahead of the best open model (DeepSeek V4 Pro 0813) at 52.9% — but you can run the open one yourself.
Scores aggregated from epoch. llmrun does not run this benchmark — see the source for methodology, or the about benchmarks for what it measures.