Cursor is an AI-native code editor built on VS Code. It uses AI models for code generation, editing, and chat. Models need strong code generation, reliable instruction-following, and fast response times for interactive editing.
Best Models for Cursor
Top 15 by tool-optimized score
Scored by: benchmark performance (90%) from MMLU, GPQA, HumanEval, SWE-bench, and 15+ standardized evaluations, with capabilities and context as tiebreakers (10%).
| # | Model | Score | Output $/M |
|---|---|---|---|
| 1 | Gemini 3.5 Flash Arena Elo: 1479 | 86 | $9.00 |
| 2 | Claude Opus 4.7 (Fast) | 85 | $150.00 |
| 3 | MiMo-V2.5-Pro Arena Elo: 1465 | 85 | $0.870 |
| 4 | GLM 5.1 Arena Elo: 1474 | 85 | $3.08 |
| 5 | Qwen3.6 Max Preview Arena Elo: 1459 | 84 | $6.24 |
| 6 | Kimi K2.6 Arena Elo: 1462 | 84 | $3.49 |
| 7 | Gemma 4 31B Arena Elo: 1452 | 84 | $0.370 |
| 8 | GLM 5 Arena Elo: 1457 | 84 | $1.92 |
| 9 | Grok 4.3 Arena Elo: 1447 | 83 | $2.50 |
| 10 | Claude Opus 4.6 (Fast) | 83 | $150.00 |
| 11 | Qwen3.6 Plus Arena Elo: 1444 | 83 | $1.95 |
| 12 | MiMo-V2-Pro Arena Elo: 1448 | 83 | $3.00 |
| 13 | GPT-5.4 Pro | 83 | $180.00 |
| 14 | Gemini 3.1 Pro Preview Custom Tools | 83 | $12.00 |
| 15 | Qwen3.5 397B A17B Arena Elo: 1445 | 83 | $2.34 |
| 16 | GPT-5.2-Codex | 83 | $14.00 |
| 17 | GLM 4.7 Arena Elo: 1443 | 83 | $1.75 |
| 18 | GPT-5.2 Pro | 83 | $168.00 |
| 19 | Claude Opus 4.1 Arena Elo: 1449 | 83 | $75.00 |
| 20 | GPT-5.5 Pro SWE-bench: 88.7% | 82 | $180.00 |
| 21 | GPT-5.5 SWE-bench: 88.7% | 82 | $30.00 |
| 22 | DeepSeek V4 Flash Arena Elo: 1433 | 82 | $0.200 |
| 23 | MiMo-V2.5 Arena Elo: 1434 | 82 | $0.280 |
| 24 | Gemma 4 26B A4B Arena Elo: 1439 | 82 | $0.330 |
| 25 | Grok 4.20 | 82 | $2.50 |
| 26 | Gemini 3.1 Flash Lite Preview Arena Elo: 1433 | 82 | $1.50 |
| 27 | GPT-5.3-Codex | 82 | $14.00 |
| 28 | GPT-5.2 Chat Arena Elo: 1476 | 82 | $14.00 |
| 29 | GPT-5 Pro | 82 | $120.00 |
| 30 | GPT-5 Codex | 82 | $10.00 |
| 31 | Claude Opus 4.7 SWE-bench: 87.6% | 81 | $25.00 |
| 32 | Qwen3.5-122B-A10B Arena Elo: 1417 | 81 | $2.08 |
| 33 | GPT-5.1-Codex-Max | 81 | $10.00 |
| 34 | GPT-5.1-Codex | 81 | $10.00 |
| 35 | GPT-5.1-Codex-Mini | 81 | $2.00 |
| 36 | o3 Deep Research | 81 | $40.00 |
| 37 | GLM 4.6 Arena Elo: 1426 | 81 | $1.74 |
| 38 | o3 Pro | 81 | $80.00 |
| 39 | MiMo-V2-Omni Arena Elo: 1414 | 80 | $2.00 |
| 40 | MiniMax M2.7 Arena Elo: 1413 | 80 | $1.20 |
| 41 | Qwen3.5-27B Arena Elo: 1408 | 80 | $1.56 |
| 42 | Qwen3 VL 235B A22B Instruct Arena Elo: 1415 | 80 | $0.880 |
| 43 | Mistral Large HumanEval: 92% | 80 | $6.00 |
| 44 | Qwen3.5-35B-A3B Arena Elo: 1396 | 79 | $1.00 |
| 45 | Qwen3.5-Flash Arena Elo: 1396 | 79 | $0.260 |
| 46 | Claude Opus 4.6 SWE-bench: 83.7% | 79 | $25.00 |
| 47 | Step 3.5 Flash Arena Elo: 1394 | 79 | $0.300 |
| 48 | DeepSeek V3.2 Exp Arena Elo: 1423 | 79 | $0.410 |
| 49 | DeepSeek V3.1 Terminus Arena Elo: 1416 | 79 | $0.950 |
| 50 | Qwen3 Next 80B A3B Instruct Arena Elo: 1402 | 79 | $1.10 |
Based on our analysis of coding benchmarks, capability matching, and pricing, Gemini 3.5 Flash currently ranks #1 for Cursor. Rankings are rebuilt as benchmark, pricing, and provider data refresh.
We score models using benchmark performance (90%) from LMArena, HumanEval, SWE-bench, MMLU, and 15+ standardized evaluations. Capabilities and context serve as tiebreakers (10%). Only models with the capabilities Cursor needs are included in the tool-specific rankings.
We currently track 331 AI models compatible with Cursor. This includes models from OpenAI, Anthropic, Google, DeepSeek, and other providers accessible via API.
Many open-source models are compatible with Cursor through API providers like OpenRouter, Together AI, and Groq. Check our rankings to see which open-source models perform best.
Rankings refresh whenever the underlying benchmark, pricing, and catalog sources refresh. That means some signals update faster than others, and the page reflects the latest verified source data available.