Cursor + DeepSeek V4 Pro
A IDE agent plus open/open-weight coding model stack for daily IDE coding with autocomplete and composer workflows.
ARscore 56.21M context$0.44/$0.87mixed stack
| Agent | Cursor |
|---|---|
| Model | DeepSeek V4 Pro |
| Agent score | 71 |
| Model SWE-bench | 79.1% |
| Combined score | 56.2 |
| Best fit | daily IDE coding with autocomplete and composer workflows |
Related Cursor stacks
Same agent, different model choices.
Cursor + Opus 4.8
ARscore 62.9, 1M context, $5/$25
view stack
Cursor + GPT-5.5
ARscore 58.6, 258K context, $6/$25
view stack
Cursor + Gemini 3.1 Pro
ARscore 57.2, 1M+ context, $1.5/$6
view stack
Cursor + Qwen 3.7 Max
ARscore 57.1, 1M context, $1.66/$4.95
view stack
Related DeepSeek V4 Pro stacks
Same model, different agent wrappers.