Back to BenchmarksFrontier
Fiction.LiveBench
Fiction comprehension and reasoning tasks
Leaderboard
(22 models)| Rank | Model | Score | Stderr |
|---|---|---|---|
| 1 | o3 | 100.00 | — |
| 2 | Grok 4 | 96.90 | — |
| 3 | GPT-5.2 | 96.90 | — |
| 4 | Gemini 2.5 Pro (Jun 2025) | 90.60 | — |
| 5 | Qwen 3 235B | 68.80 | — |
| 6 | chatgpt-4o-01-29-2025 | 65.60 | — |
| 7 | Grok-3 mini | 65.60 | — |
| 8 | GPT-4.1 | 63.90 | — |
| 9 | o4-mini-2025-04-16 medium | 62.50 | — |
| 10 | Minimax M2 | 59.40 | — |
| 11 | o1 | 53.10 | — |
| 12 | DeepSeek V3 | 53.10 | — |
| 13 | Claude 3.7 Sonnet | 53.10 | — |
| 14 | GPT-4.1 mini | 46.90 | — |
| 15 | Qwen3-235B-A22B | 44.40 | — |
| 16 | Kimi K2 Instruct | 40.60 | — |
| 17 | Claude Opus 4.5 | 37.50 | — |
| 18 | Gemini 2.0 Pro Exp (Feb 2025) | 37.50 | — |
| 19 | Gemini 2.0 Flash Thinking Exp | 37.50 | — |
| 20 | Llama-4-Maverick-17B-128E-Instruct | 36.40 | — |
| 21 | DeepSeek R1 | 33.30 | — |
| 22 | Llama 4 Scout | 27.30 | — |
Data source: Epoch AI, “Data on AI Benchmarking”. Published at epoch.ai
Licensed under CC-BY 4.0