AI model leaderboard
How today's AI models do on real questions, not benchmarks: a rating from blind head-to-head votes (0 so far), median time to first word and reliability from the last 30 days, and price per answer.
| # | Model | Maker | Rating | Win rate | Time to first word | Reliability | Credits per answer |
|---|---|---|---|---|---|---|---|
| — | Claude Sonnet 5.5 | Anthropic | — | — | 5.0 s | 100.0% | 22 |