Leaderboard
SWE-Bench Pro
56 models scored · metric: resolve 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.Laguna XS 2.1635
- 2.GPT-5.6 Luna279
- 3.Step 3.7 Flash135
- 4.GPT-5.4 nano128
- 5.LongCat-2.0113
| # | Model | Score | Best in /M | Best out /M | Pts / $ | Context |
|---|---|---|---|---|---|---|
| 🥇 | Claude Fable 5anthropic | 80.3 | $10 | $50 | 4.0 | 1M |
| 🥈 | Claude Opus 5anthropic | 79.2 | $5 | $25 | 7.9 | 1M |
| 🥉 | Claude Opus 4.8anthropic | 69.2 | $5 | $25 | 6.9 | 1M |
| 4 | Grok 4.5xai | 64.7 | $2 | $6 | 21.6 | 500K |
| 5 | GPT-5.6 Solopenai | 64.6 | $2.50 | $15 | 11.5 | 1.05M |
| 6 | Claude Opus 4.7anthropic | 64.3 | $5 | $25 | 6.4 | 1M |
| 7 | GPT-5.6 Terraopenai | 63.4 | $1 | $6 | 28.2 | 1.05M |
| 8 | Claude Sonnet 5anthropic | 63.2 | $2 | $10 | 15.8 | 1M |
| 9 | GPT-5.6 Lunaopenai | 62.7 | $0.1 | $0.6 | 279 | 1.05M |
| 10 | GLM-5.2zhipuai | 62.1 | $1.40 | $4.40 | 28.9 | 1M |
| 11 | Muse Spark 1.1meta | 61.5 | $1.25 | $4.25 | 30.8 | 1M |
| 12 | Qwen3.7 Maxalibaba | 60.6 | $2.50 | $7.50 | 16.2 | 1M |
| 13 | LongCat-2.0meituan | 59.5 | $0.3 | $1.20 | 113 | 1M |
| 14 | GPT-5.4openai | 59.1 | $2.20 | $14 | 11.5 | 1.05M |
| 15 | MiniMax-M3minimax | 59 | $0.3 | $1.20 | 112 | 512K |
| 16 | Gemini 3.6 Flashgoogle | 58.7 | $0.75 | $3.75 | 39.1 | 1.05M |
| 17 | GPT-5.5openai | 58.6 | $4.55 | $27.27 | 5.7 | 1.05M |
| 18 | MiMo-V2.5-Proxiaomi | 57.2 | $0.435 | $0.87 | 105 | 1.05M |
| 19 | Step 3.7 Flashstepfun | 56.3 | $0.185 | $1.11 | 135 | 256K |
| 20 | MiniMax-M2.7minimax | 56.2 | $0.3 | $1.20 | 107 | 205K |
| 21 | Gemini 3.5 Flashgoogle | 55.1 | $1.50 | $9 | 16.3 | 1.05M |
| 22 | GPT-5.4 miniopenai | 54.4 | $0.68 | $4 | 36.0 | 400K |
| 23 | Gemini 3.1 Pro Previewgoogle | 54.2 | $2 | $12 | 12.0 | 1.05M |
| 24 | Gemini 3.5 Flash Litegoogle | 54.2 | $0.3 | $2.50 | 63.8 | 1.05M |
| 25 | GPT-5.4 nanoopenai | 52.4 | $0.18 | $1.10 | 128 | 400K |
| 26 | Claude Opus 4.6anthropic | 51.9 | $5 | $25 | 5.2 | 1M |
| 27 | Muse Glimmer 30Bmeta | 51.2 | $0.3 | $1.10 | 102 | 131K |
| 28 | Laguna XS 2.1poolside | 47.6 | $0.06 | $0.12 | 635 | 262K |
| 29 | Gemini 3.1 Pro Previewgoogle | 46.1 | $2 | $12 | 10.2 | 1.05M |
| 30 | Claude Opus 4.5anthropic | 45.89 | $5 | $25 | 4.6 | 200K |
| 31 | Claude Opus 4.7anthropic | 44.9 | $5 | $25 | 4.5 | 1M |
| 32 | Claude Sonnet 4.5 (latest)anthropic | 43.6 | $3 | $15 | 7.3 | 200K |
| 33 | Gemini 3 Pro Previewgoogle | 43.3 | $2 | $12 | 9.6 | 1.05M |
| 34 | GPT-5openai | 41.78 | $1.10 | $9 | 13.6 | 400K |
| 35 | GPT-5.2 Codexopenai | 41.04 | $1.60 | $13 | 9.2 | 400K |
| 36 | North Mini Codecohere | 40.2 | — | — | — | 256K |
| 37 | Claude Haiku 4.5 (latest)anthropic | 39.45 | $1 | $5 | 19.7 | 200K |
| 38 | Qwen3-Coder 480B-A35B Instructalibaba | 38.7 | $1.50 | $7.50 | 12.9 | 262K |
| 39 | MiniMax-M2.1minimax | 36.81 | $0.3 | $1.20 | 70.1 | 205K |
| 40 | Claude Opus 4.7anthropic | 36.4 | $5 | $25 | 3.6 | 1M |
| 41 | Gemini 3 Flash Previewgoogle | 34.63 | $0.5 | $3 | 30.8 | 1.05M |
| 42 | Claude Opus 4.7anthropic | 34.4 | $5 | $25 | 3.4 | 1M |
| 43 | GPT-5.5openai | 30.9 | $4.55 | $27.27 | 3.0 | 1.05M |
| 44 | GPT-5.2openai | 29.94 | $1.60 | $13 | 6.7 | 400K |
| 45 | Kimi K2.6moonshotai | 27.3 | $0.5 | $2.60 | 26.6 | 262K |
| 46 | GPT-5.5openai | 26.2 | $4.55 | $27.27 | 2.6 | 1.05M |
| 47 | GPT-5.5openai | 24.9 | $4.55 | $27.27 | 2.4 | 1.05M |
| 48 | Qwen3 235B-A22Balibaba | 21.41 | $0.7 | $2.80 | 17.5 | 131K |
| 49 | GLM-5.1zhipuai | 19.8 | $1.40 | $4.40 | 9.2 | 200K |
| 50 | GPT-5.4openai | 18.9 | $2.20 | $14 | 3.7 | 1.05M |
| 51 | GPT-5.4openai | 18.4 | $2.20 | $14 | 3.6 | 1.05M |
| 52 | DeepSeek V4 Prodeepseek | 18 | $0.435 | $0.87 | 33.1 | 1M |
| 53 | Gemini 3.1 Pro Previewgoogle | 15.1 | $2 | $12 | 3.4 | 1.05M |
| 54 | Claude Sonnet 4.6anthropic | 14.9 | $3 | $15 | 2.5 | 1M |
| 55 | Claude Opus 4.6anthropic | 11.8 | $5 | $25 | 1.2 | 1M |
| 56 | GLM-4.6zhipuai | 9.67 | $0.6 | $2.20 | 9.7 | 205K |
Cheapest scorer: Laguna XS 2.1 at $0.06 input /M.