HoWToBench - Contract: leaderboard

HoWToBench (2026) tests long-form Chinese writing against curated human references; GPT-4o rates content traits 1 to 10 and format traits 0, 5 or 10, and the Tree-of-Writing method combines them with explicit per-instruction weights. Mean score on the contract genre instructions; higher is better.

Metric: Tree-of-Writing score: GPT-4o rates each content trait 1-10 against the human reference and format traits 0, 5 or 10, combined with per-instruction LLM-assigned weights that sum to 1, mean over the contract writing instructions of HoWToBench; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1O3 Mini (2025-01-31)5.8#502
2HoWToBench DeepSeek-R1 (checkpoint unspecified)5.52
3HoWToBench DeepSeek-V3 (checkpoint unspecified)5.33#669
4Doubao-pro-2412255.24#817
5GLM-4-Plus-2501115.18#836
6Claude 3.5 Sonnet (20241022)5.11#651
7Gemini 2.0 Flash5.08#731
8Claude 3 Haiku (20240307)5.06#2415
9GPT-4o (2024-11-20)4.97#860
10Llama 3.3 70B Instruct4.37#1394

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=howtobench-contract · How It Works · Data refreshed daily, snapshot 2026-11-22.