Fiction Eval leaderboard · comparisons
All model comparisons
From the July 2026 run — every model judged blind against every other, head-to-head.
- Claude Opus 4.8 vs Claude Sonnet 4.6
- Claude Opus 4.8 vs Grok 4.3
- Claude Opus 4.8 vs GLM 5.2
- Claude Sonnet 4.6 vs Grok 4.3
- Claude Opus 4.8 vs Gemini 3.1 Pro
- Claude Sonnet 4.6 vs GLM 5.2
- Claude Opus 4.8 vs DeepSeek V3.2
- Claude Sonnet 4.6 vs Gemini 3.1 Pro
- Grok 4.3 vs GLM 5.2
- Claude Opus 4.8 vs Mistral Large
- Claude Sonnet 4.6 vs DeepSeek V3.2
- Grok 4.3 vs Gemini 3.1 Pro
- Claude Opus 4.8 vs Llama 4 Maverick
- Claude Sonnet 4.6 vs Mistral Large
- GLM 5.2 vs Gemini 3.1 Pro
- Grok 4.3 vs DeepSeek V3.2
- Claude Sonnet 4.6 vs Llama 4 Maverick
- GLM 5.2 vs DeepSeek V3.2
- Grok 4.3 vs Mistral Large
- Gemini 3.1 Pro vs DeepSeek V3.2
- GLM 5.2 vs Mistral Large
- Grok 4.3 vs Llama 4 Maverick
- Gemini 3.1 Pro vs Mistral Large
- GLM 5.2 vs Llama 4 Maverick
- DeepSeek V3.2 vs Mistral Large
- Gemini 3.1 Pro vs Llama 4 Maverick
- DeepSeek V3.2 vs Llama 4 Maverick
- Mistral Large vs Llama 4 Maverick
How it works: every pair of models is judged blind on the same scene, with the passages'
order flipped so being shown first can't sway it, by GPT-5.4 — a family that isn't on the
board, so nobody scores their own side. Each score carries a likely range; overlapping ranges
are called a tie. Slop is scored separately by a fixed checklist, not an AI. Full board,
every prompt, and the FAQ: thebookfactoryai.com/board.
Get each new board by email on the model-drop list.
Writing a book of your own? Book Factory runs the same craft checks on full manuscripts.
Writing a book of your own? Book Factory runs the same craft checks on full manuscripts.