Creative Writing (Lechmazur): leaderboard
Evaluates creative fiction writing quality across 30+ models using structured prompts and LLM-as-judge scoring. Measures prose quality, creativity, coherence, and adherence to constraints.
Metric: Mean Score. Source: github.com. Saturation forecast: Around 2029. 59 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5.1 (High) | 3.76 | #9 (Claude Fable 5.1) |
| 2 | Claude Opus 5 (xHigh) | 3.73 | #12 (Claude Opus 5) |
| 3 | Claude Opus 5.5 (High) | 3.72 | #3 (Claude Opus 5.5) |
| 4 | GPT-6 Astra (High) | 3.44 | #6 (GPT-6 Astra) |
| 5 | Claude Opus 5 (High) | 3.37 | #12 (Claude Opus 5) |
| 6 | GLM-5.3 (Max) | 3.03 | #39 (GLM-5.3) |
| 7 | Claude Fable 5 (High) | 2.85 | #16 (Claude Fable 5) |
| 8 | GPT-5.6 Sol (xHigh) | 2.47 | #19 (GPT-5.6 Sol) |
| 9 | GPT-5.5 (xHigh) | 2.43 | #31 (GPT-5.5) |
| 10 | Kimi K3 | 2.42 | #30 |
| 11 | GPT-5.6 Sol (High) | 2.28 | #19 (GPT-5.6 Sol) |
| 12 | GPT-5.4 (Medium) | 1.97 | #79 (GPT-5.4) |
| 13 | GPT-5.4 (xHigh) | 1.96 | #79 (GPT-5.4) |
| 14 | Claude Opus 4.7 (High) | 1.76 | #49 (Claude Opus 4.7) |
| 15 | Gemini 4 Argon (High) | 0.8 | #8 (Gemini 4 Argon) |
No result here: #7 GPT-6.1 Sol, #10 Muse Wonder Full.
Interactive version: theaggregate.ai/benchmark?slug=creative-writing-lechmazur · How It Works · Data refreshed daily, snapshot 2026-10-20.