vercel/
The Open Framework for Building Agents
Run by mupt-ai · Released 2026-10-02 · 30 tasks · 5 settings
Which coding agent works best on vercel/eve? Most accurate: GPT-6.1 Sol, 76.7% at $0.74 per task. Best for less: DeepSeek V4.1 Flash, 66.7% at $0.35. 5 model settings scored on 30 tasks from its merged pull requests.
All Settings
| Model | Harness | Reasoning | Access | Accuracy | Cost / Task | Frontier |
|---|---|---|---|---|---|---|
| GPT-6.1 Sol | Codex | High | API Key | 76.7% | $0.74 | Yes |
| Kimi K3 | Pi | High | Vercel AI Gateway | 73.3% | $1.89 | No |
| DeepSeek V4.1 Flash | Pi | High | Vercel AI Gateway | 66.7% | $0.35 | Yes |
| GPT-6 Luna | Codex | High | API Key | 60% | $0.031 | Yes |
| GLM 5.3 Flash | Pi | High | Vercel AI Gateway | 56.7% | $0.20 | No |