Leaderboard
AutomationBench
5 models scored · metric: success rate. Prices are the cheapest listed across all serving providers. Pts/$ = score ÷ blended price (3×input + output) / 4 — how much score each dollar buys.
Best value · top 5 by pts per dollar
- 1.DeepSeek V4 Flash 0731239
- 2.Gemini 3.7 Flash40.5
- 3.Kimi K35.9
- 4.Claude Opus 52.6
- 5.Claude Fable 50.9
| # | Model | Score | Best in /M | Best out /M | Pts / $ | Context |
|---|---|---|---|---|---|---|
| 🥇 | Kimi K3moonshotai | 30.8 | $2.50 | $13.50 | 5.9 | 1.05M |
| 🥈 | Gemini 3.7 Flashgoogle | 30.4 | $0.375 | $1.88 | 40.5 | 1.05M |
| 🥉 | Claude Opus 5anthropic | 26 | $5 | $25 | 2.6 | 1M |
| 4 | DeepSeek V4 Flash 0731deepseek | 25.1 | $0.08 | $0.18 | 239 | 1M |
| 5 | Claude Fable 5anthropic | 17.4 | $10 | $50 | 0.9 | 1M |
Cheapest scorer: DeepSeek V4 Flash 0731 at $0.08 input /M.