SimpleQA leaderboard
As of November 14, 2026, among the results listed here, Inference.net Schematron V2 Small has the highest reported SimpleQA score, 83.1, published by Inference.net on April 16, 2026, ahead of Inference.net Schematron V2 Turbo at 79.4. 7 models have a SimpleQA result in our records.
Scores as the model makers published them, not re-run by us; each row links to its announcement.
Best reported score
83.1
Inference.net Schematron V2 Small
Reported by Inference.net, April 16, 2026
7 models ranked · 3 makers
| Rank | Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|---|
| 1 | Schematron V2 Small | 83.1 | Not published | Inference.net | April 16, 2026 |
| 2 | Schematron V2 Turbo | 79.4 | Not published | Inference.net | April 16, 2026 |
| 3 | Kimi-K2-Base | 35.3 | Base model evaluation results; Correct | Moonshot AI | July 11, 2025 (model release) |
| 4 | Kimi K2 0711 | 31 | Instruction model evaluation results; Correct | Moonshot AI | July 11, 2025 (model release) |
| 5 | o4 MiniRetired | 0.2 | accuracy | OpenAI | August 5, 2025 (model release) |
| 6 | gpt-oss-120b | 0.2 | accuracy | OpenAI | August 5, 2025 (model release) |
| 7 | gpt-oss-20b | 0.1 | accuracy | OpenAI | August 5, 2025 (model release) |
How much to trust these numbers
- Source
- Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
- Settings
- Settings differ between makers. Check the Setting column before comparing two scores.
- Spread
- 83.1 to 0.1 across 7 models.
- Freshness
- Newest result published April 16, 2026; oldest July 11, 2025.
What SimpleQA measures
SimpleQA tests how accurately a model answers short, fact-seeking questions, with higher scores indicating better factual accuracy. Source: Measuring short-form factuality in large language models