ELO ratings from LMSYS Chatbot Arena as reported in the LLM Benchmark Matrix. Based on crowdsourced pairwise human preference judgments across diverse prompts.
This definition has not completed editorial and provenance review. Verify its metric, version, source, and evaluation configuration upstream before using it for a model-selection decision.
Gold standard for human preference evaluation. Large-scale crowdsourced comparisons.
ELO scale (not 0–100%). Matrix snapshot may lag live Arena rankings. We also have dedicated Arena scraper data.
Higher is better. ELO rating (typical range 1338–1502 for frontier models).
ELO scale (not 0–100%).
| # | Model | Score |
|---|---|---|
| 1 | 1502.0 | |
| 2 | 1502.0 | |
| 3 | 1501.0 | |
| 4 | 1490.0 | |
| 5 | 1490.0 | |
| 6 | 1483.0 | |
| 7 | 1483.0 | |
| 8 | 1473.0 | |
| 9 | 1473.0 | |
| 10 | 1468.0 | |
| 11 | 1468.0 | |
| 12 | 1465.0 | |
| 13 | 1464.0 | |
| 14 | 1464.0 | |
| 15 | 1460.0 | |
| 16 | 1460.0 | |
| 17 | 1460.0 | |
| 18 | 1460.0 | |
| 19 | 1445.0 | |
| 20 | 1445.0 | |
| 21 | 1440.0 | |
| 22 | 1440.0 | |
| 23 | 1440.0 | |
| 24 | 1437.0 | |
| 25 | 1420.0 | |
| 26 | 1420.0 | |
| 27 | 1418.0 | |
| 28 | 1417.0 | |
| 29 | 1410.0 | |
| 30 | 1410.0 | |
| 31 | 1408.0 | |
| 32 | 1402.0 | |
| 33 | 1402.0 | |
| 34 | 1382.0 | |
| 35 | 1382.0 | |
| 36 | 1372.0 | |
| 37 | 1372.0 | |
| 38 | 1364.0 | |
| 39 | 1364.0 | |
| 40 | 1350.0 | |
| 41 | 1338.0 |