Leaderboard
Preference ratings from blind human votes, on an Elo-like scale — cost and speed were never visible to voters.
Claude Opus 5
1623
±313
1/2 picked · $1.26/gen
provisional
Claude Fable 5
1557
±96
20/60 picked · $1.04/gen
GPT 5.6 Sol
1486
±97
18/72 picked · $0.305/gen
| # | Model | Rating | W / L | Votes | Avg cost | Median time |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5Anthropicprovisional | 1623±313 | 2 | $1.26 | ||
| 2 | Claude Fable 5Anthropic | 1557±96 | 60 | $1.04 | ||
| 3 | GPT 5.6 SolOpenAI | 1486±97 | 72 | $0.305 | ||
| 4 | Kimi K3Moonshot | 1481±101 | 66 | $0.424 | ||
| 5 | GLM 5.2Z.ai | 1432±104 | 73 | $0.066 |
Ratings come from one preference model fitted to every blind vote at once (Plackett–Luce, on an Elo-like scale where +400 ≈ 10:1 odds). Because it’s fitted globally rather than accumulated vote by vote, a model added later is judged on its own record — not penalised for arriving late. ± is the 95% confidence interval; overlapping intervals mean the gap isn’t established yet. “Provisional” means too few votes to place confidently. W/L is times picked vs times shown. Cost and time are per full generation. ▲/▼ is the 7-day rank move.
Rating vs cost
Up and left is the value frontier — preferred by humans, cheap to run. The leader is highlighted in gold.
Bubble size reflects vote volume.