StaminaBench (OpenCode): leaderboard
Metric: OpenCode harness: average consecutive interaction turns passed before the first failing change request, over 20 procedurally generated REST-API scenarios of up to 101 turns with two retries; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GLM-5 | 57 |
| 2 | Qwen 3.5 122B A10B | 39.4 |
| 3 | Kimi K2.5 | 26.8 |
| 4 | Qwen3 Coder Next | 7.6 |
| 5 | Devstral Small 2 | 5.5 |
| 6 | Devstral 2 | 4.8 |
Interactive version: theaggregate.ai/benchmark?slug=staminabench-opencode · How It Works · Data refreshed daily, snapshot 2026-09-29.