Leaderboard
SWE-Bench Verified
36 models scored · metric: resolved. 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.Laguna XS 2.1945
- 2.DeepSeek V4 Flash780
- 3.Hy3681
- 4.Step 3.5 Flash522
- 5.Qwen3.6 35B-A3B280
| # | Model | Score | Best in /M | Best out /M | Pts / $ | Context |
|---|---|---|---|---|---|---|
| 🥇 | Claude Opus 5anthropic | 96 | $5 | $25 | 9.6 | 1M |
| 🥈 | Claude Fable 5anthropic | 95 | $10 | $50 | 4.8 | 1M |
| 🥉 | Claude Opus 4.8anthropic | 88.6 | $5 | $25 | 8.9 | 1M |
| 4 | Claude Sonnet 5anthropic | 85.2 | $2 | $10 | 21.3 | 1M |
| 5 | DeepSeek V4 Prodeepseek | 80.6 | $0.435 | $0.87 | 148 | 1M |
| 6 | MiniMax-M3minimax | 80.5 | $0.3 | $1.20 | 153 | 512K |
| 7 | Qwen3.7 Maxalibaba | 80.4 | $2.50 | $7.50 | 21.4 | 1M |
| 8 | Kimi K2.6moonshotai | 80.2 | $0.5 | $2.60 | 78.2 | 262K |
| 9 | MiniMax-M2.7minimax | 79.9 | $0.3 | $1.20 | 152 | 205K |
| 10 | DeepSeek V4 Flashdeepseek | 79 | $0.081 | $0.162 | 780 | 1M |
| 11 | MiMo-V2.5-Proxiaomi | 78.9 | $0.435 | $0.87 | 145 | 1.05M |
| 12 | Hy3tencent | 78 | $0.066 | $0.26 | 681 | 256K |
| 13 | Mistral Medium 3.5mistral | 77.6 | $1.50 | $7.50 | 25.9 | 262K |
| 14 | Mistral Medium (latest)mistral | 77.6 | $0.4 | $2 | 97.0 | 262K |
| 15 | Qwen3.6 27Balibaba | 77.2 | $0.203 | $2.24 | 108 | 262K |
| 16 | Step 3.7 Flashstepfun | 76.5 | $0.185 | $1.11 | 184 | 256K |
| 17 | Qwen3.5 397B-A17Balibaba | 76.4 | $0.6 | $3.60 | 56.6 | 262K |
| 18 | Muse Glimmer 30Bmeta | 76 | $0.3 | $1.10 | 152 | 131K |
| 19 | MiniMax-M2.5minimax | 75.8 | $0.3 | $1.20 | 144 | 205K |
| 20 | Step 3.5 Flashstepfun | 74.4 | $0.09 | $0.3 | 522 | 256K |
| 21 | Hy3 previewtencent | 74.4 | $0.172 | $0.572 | 274 | 256K |
| 22 | MiniMax-M2.1minimax | 74 | $0.3 | $1.20 | 141 | 205K |
| 23 | GLM-4.7zhipuai | 73.8 | $0.6 | $2.20 | 73.8 | 205K |
| 24 | Qwen3.6 35B-A3Balibaba | 73.4 | $0.19 | $0.48 | 280 | 262K |
| 25 | GLM-5zhipuai | 72.8 | $1 | $3.20 | 47.0 | 205K |
| 26 | Qwen3.5 27Balibaba | 72.4 | $0.3 | $2.40 | 87.8 | 262K |
| 27 | Qwen3.5 122B-A10Balibaba | 72 | $0.4 | $3.20 | 65.5 | 262K |
| 28 | Kimi K2 Thinkingmoonshotai | 71.3 | $0.47 | $2 | 83.6 | 262K |
| 29 | Laguna XS 2.1poolside | 70.9 | $0.06 | $0.12 | 945 | 262K |
| 30 | Kimi K2.5moonshotai | 70.8 | $0.3 | $1.90 | 101 | 262K |
| 31 | Nemotron 3 Ultra 550B A55Bnvidia | 70.7 | $0.5 | $2.20 | 76.4 | 1M |
| 32 | MiniMax-M2minimax | 69.4 | $0.3 | $1.20 | 132 | 197K |
| 33 | North Mini Codecohere | 67.6 | — | — | — | 256K |
| 34 | Devstral Mediummistral | 61.6 | — | — | — | 128K |
| 35 | GLM-4.7-Flashzhipuai | 59.2 | — | — | — | 200K |
| 36 | Devstral Smallmistral | 53.6 | — | — | — | 128K |
Cheapest scorer: Laguna XS 2.1 at $0.06 input /M.