Leaderboard
Agents' Last Exam
4 models scored · metric: score. 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 4 by pts per dollar
- 1.DeepSeek V4 Flash 0731240
- 2.GPT-5.6 Luna224
- 3.GPT-5.6 Terra22.4
- 4.GPT-5.6 Sol9.4
| # | Model | Score | Best in /M | Best out /M | Pts / $ | Context |
|---|---|---|---|---|---|---|
| 🥇 | GPT-5.6 Solopenai | 52.7 | $2.50 | $15 | 9.4 | 1.05M |
| 🥈 | GPT-5.6 Terraopenai | 50.4 | $1 | $6 | 22.4 | 1.05M |
| 🥉 | GPT-5.6 Lunaopenai | 50.3 | $0.1 | $0.6 | 224 | 1.05M |
| 4 | DeepSeek V4 Flash 0731deepseek | 25.2 | $0.08 | $0.18 | 240 | 1M |
Cheapest scorer: DeepSeek V4 Flash 0731 at $0.08 input /M.