Leaderboard
MLE-Bench
2 models scored · metric: average position 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 2 by pts per dollar
- 1.Gemini 3.5 Flash Lite46.1
- 2.Gemini 3.6 Flash42.6
| # | Model | Score | Best in /M | Best out /M | Pts / $ | Context |
|---|---|---|---|---|---|---|
| 🥇 | Gemini 3.6 Flashgoogle | 63.9 | $0.75 | $3.75 | 42.6 | 1.05M |
| 🥈 | Gemini 3.5 Flash Litegoogle | 39.2 | $0.3 | $2.50 | 46.1 | 1.05M |
Cheapest scorer: Gemini 3.5 Flash Lite at $0.3 input /M.