Benchmark
Arena ELO - Exclude Ties
People compare model answers and choose which they prefer. Higher scores are better.
All human preference benchmarksScores from LMArena data.
- Models measured
- 366
- Scores recorded
- 423
- Sources
- 1
- IQ weight
- 1.28
Score distribution
- 1,247.7
- Mean
- 1,316.3
- Median
- 217.2
- Std dev
- 908.5
- 10th percentile
- 1,465.2
- 90th percentile
- 533.7
- Min
This benchmark has a weight of 1.28 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 401 | Vicuna 7B· LMArena data | 841 |
| 402 | StripedHyena Nous 7B· LMArena data | 839.9 |
| 403 | PaLM 2· LMArena data | 830.1 |
| 404 | OLMo 7B Instruct· LMArena data | 827.3 |
| 405 | Llama 3.2 1B Instruct· LMArena data | 822.9 |
| 406 | Mistral 7B Instruct· LMArena data | 819 |
| 407 | SmolLM2 1.7B Instruct· LMArena data | 797 |
| 408 | Gemma 1.1 2B IT· LMArena data | 796.4 |
| 409 | Koala 13B· LMArena data | 782.1 |
| 410 | Gemma 2B IT· LMArena data | 779.7 |
| 411 | Qwen1.5-4B-Chat· LMArena data | 763.1 |
| 412 | ChatGLM3 6B· LMArena data | 741.6 |
| 413 | GPT4All 13B Snoozy· LMArena data | 727.2 |
| 414 | MPT 7B Chat· LMArena data | 726.9 |
| 415 | RWKV-4 Raven 14B· LMArena data | 721.1 |
| 416 | Alpaca 13B· LMArena data | 697.4 |
| 417 | ChatGLM2 6B· LMArena data | 684.8 |
| 418 | OpenAssistant Pythia 12B· LMArena data | 671.6 |
| 419 | ChatGLM 6B· LMArena data | 661.7 |
| 420 | FastChat-T5 3B· LMArena data | 635 |
| 421 | StableLM Tuned Alpha 7B· LMArena data | 583.5 |
| 422 | Dolly 2.0 12B· LMArena data | 567.8 |
| 423 | LLaMA 13B· LMArena data | 533.7 |
Showing 401–423 of 423 results.
Data credits: LMArena, CC BY 4.0 (licence)