Benchmarks
Human preference benchmarks
People compare model answers and choose which they prefer.
We track 136 benchmarks in this topic. Main benchmarks are listed by how many models they cover, most first. Where a benchmark has separate results for a subset of its tasks, those results appear right under it. Click one for its leaderboard.
Showing 1–100 of 136 benchmarks.
- Arena ELO - English366 models
- Arena ELO - Exclude Ties366 models
- Arena ELO - Hard Prompts366 models
- Arena ELO - Non English366 models
- Arena ELO - Creative Writing364 models
- Arena ELO - Hard Prompts English364 models
- Arena ELO - Multi Turn364 models
- Arena ELO - Industry Mathematical348 models
- Arena ELO - Longer Query345 models
- Arena ELO - Chinese340 models
- Arena ELO - Russian332 models
- Arena ELO - Expert318 models
- Arena ELO - German273 models
- Arena ELO - French262 models
- Arena ELO - Spanish259 models
- Arena ELO - Japanese245 models
- Arena ELO - Korean244 models
- Arena ELO - Polish207 models
- Arena ELO Text Factuality170 models
- Arena ELO Text Factuality - Exclude Ties169 models
- Arena ELO Text Factuality - Hard Prompts168 models
- Arena ELO Text Factuality - Non English168 models
- Arena ELO Text Factuality - English167 models
- Arena ELO Text Factuality - Longer Query160 models
- Design Arena Elo (models/gamedev)148 models
- Design Arena Elo (models/website)148 models
- Arena ELO Vision147 models
- Arena ELO Vision - English147 models
- Arena ELO Vision Style-Controlled147 models
- Arena ELO Text Factuality - Multi Turn142 models
- Design Arena Elo (models/uicomponent)141 models
- Design Arena Elo (models/3d)135 models
- Arena ELO Text Factuality - Russian130 models
- Arena ELO Vision - Chinese118 models
- Arena ELO Text Factuality - Expert114 models
- Design Arena Elo (models/dataviz)114 models
- Arena ELO Text Factuality - Chinese108 models
- Arena ELO Vision - Diagram108 models
- Arena ELO Vision - Homework104 models
- Design Arena Elo (models/svg)100 models
- Arena ELO Vision - Humor90 models
- Arena ELO Text Factuality - Spanish86 models
- Arena ELO Text Factuality - French82 models
- Design Arena Elo (models/ascii)76 models
- Arena ELO Image Edit58 models
- Design Arena Elo (models/image)58 models
- Arena ELO Text Factuality - Polish57 models
- Design Arena Elo (models/logo)50 models
- Arena ELO Text Factuality - German49 models
- Arena ELO Text Factuality - Japanese47 models
- Arena ELO Text Factuality - Korean46 models
- Arena ELO Search44 models
- Arena ELO Search Style-Controlled44 models
- Arena ELO Search Factuality42 models
- Arena ELO Document40 models
- Arena ELO Document Style-Controlled40 models
- Arena ELO Vision - Creative Writing32 models