Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

WritingBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
WritingWritingBench
Score85.87
104
WritingWritingBench v1 (test)
Average Score85.3
61
Creative WritingWritingBench
Score80.39
42
Instruction FollowingWritingBench
Average Score81
29
Instruction FollowingWritingBench (Out-of-Domain)
Average Score7.9
23
Open-ended WritingWritingBench
Score75.76
20
Reward AggregationWritingBench
Score85
18
Long-form WritingWritingBench
Score88.27
18
Controllable writingWritingBench (WB)
WB-A Score79.8
17
Long-form Writing EvaluationWritingBench In-domain (test)
Elo Rating1,574
15
Long-form GenerationWritingBench length-constrained
L_R Score8.6
14
Long-form writingWritingBench D4 English literature/arts
Score (D4 Lit/Arts)9.5
14
Writing capabilitiesWritingBench (test)
Score8.56
12
Writing capability evaluationWritingBench November 2025 (official leaderboard)
Overall Score83.87
9
Long-form generationWritingBench
Score5.1
6
Writing and Arena EvaluationWritingBench
Accuracy87.63
3
Generative PerformanceWritingBench
Pearson r0.62
1
Showing 17 of 17 rows