Hemingway-bench
Chat & Writing · 2026-02-04
Elo rating on real-world writing tasks graded by expert writers.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 1110 |
| Gemini 3.7 Flash | 1105 |
| Claude Fable 5.1 | 1095 |
| Gemini 3.8 Flash | 1093 |
| Gemini 3.6 Flash | 1075 |
| Gemini 3.1 Pro Preview | 1073 |
| Gemini 3.5 Flash | 1071 |
| Kimi K3 | 1069 |
| GPT-5.6 Sol | 1057 |
| Gemini 3 Pro Preview | 1056 |
| GLM-5.3 | 1053 |
| Gemini 3 Flash Preview | 1052 |
| Claude Opus 5 | 1047 |
| Opus 4.6 | 1045 |
| Opus 4.8 | 1044 |
| Opus 4.7 | 1042 |
| GLM-5.2 | 1036 |
| GPT-5.5 | 1035 |
| GLM-5.3 Flash | 1034 |
| Kimi K2.6 | 1021 |
| Opus 4.5 | 1016 |
| Muse Spark 1.2 | 1013 |
| Sonnet 4.6 | 1001 |
| DeepSeek-V4-Pro | 999 |
| GPT-5.2 Instant (source unspecified) | 994 |
| Gemini 3.5 Flash-Lite | 990 |
| Qwen3.5 Plus (2026-02-15) | 990 |
| Kimi K2.5 | 988 |
| GPT-5.4 | 985 |
| Muse Spark 1.1 | 985 |
| Grok 4.6 | 983 |
| DeepSeek V4 Flash Vision Exp | 981 |
| Grok 4.5 | 972 |
| Hy3 | 971 |
| DeepSeek-V4-Flash | 969 |
| GPT-5.2 | 954 |
| Qwen3 Max (rolling alias) | 948 |
| Grok 4.1 Fast | 916 |
| Muse Glimmer 30B | 913 |
| Nemotron 3.5 Lightning 30B A3B (precision unspecified) | 888 |