legal benchmark
WritingBench Leaderboard
WritingBench evaluates large language models on generative writing tasks across 6 core writing domains and 100 subdomains using 1,239 queries.
Updated Sep 5, 2026
Models15
Model coverage15
MetricScore
EvidenceB
WritingBench Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score88.30% | Percentile100.00% | Participants15 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score87.30% | Percentile92.86% | Participants15 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score86.70% | Percentile85.71% | Participants15 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score86.20% | Percentile78.57% | Participants15 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score85.50% | Percentile71.43% | Participants15 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score85.50% | Percentile64.29% | Participants15 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score85.20% | Percentile57.14% | Participants15 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score85.20% | Percentile50.00% | Participants15 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score84.60% | Percentile42.86% | Participants15 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score84.00% | Percentile35.71% | Participants15 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score83.10% | Percentile28.57% | Participants15 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score82.90% | Percentile21.43% | Participants15 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score82.60% | Percentile14.29% | Participants15 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score82.50% | Percentile7.14% | Participants15 | EvidenceC | Evaluated |
| Rank15 | ModelMA | Score73.80% | Percentile0.00% | Participants15 | EvidenceC | Evaluated |
WritingBench Highlights
The leading models and scores on this benchmark.
WritingBench Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for WritingBench
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is WritingBench?
What WritingBench measures and how its scores work.
WritingBench is a benchmark for evaluating large language models' generative writing capabilities across Academic & Engineering, Finance & Business, Politics & Law, Literature & Art, Education, and Advertising & Marketing.
It reports a Score as a ratio based on query-dependent criteria generated for each task and evaluates responses on style, format, and length dimensions.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- WritingBench
- Modality
- text
- Primary category
- legal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about WritingBench.
Which model scores highest on WritingBench?
Qwen3-235B-A22B-Thinking-2507 is currently ranked first with 88.30%.
What are the top three models on WritingBench?
The current leaders are Qwen3-235B-A22B-Thinking-2507 (88.30%), Qwen3-Next-80B-A3B-Instruct (87.30%), and Qwen3 VL 235B A22B Thinking (86.70%).
Which WritingBench model has the lowest official input price?
Qwen3-Next-80B-A3B-Instruct has the lowest matched official input price at $0.50 input / $2.0 output per 1M tokens.
Which models are fastest among WritingBench results?
The fastest matched records are Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.