TeamBench - Full Team: leaderboard
Metric: Pass rate (%) on the 90-task stratified TeamBench leaderboard subset (operating-system-enforced role separation in Docker sandboxes; a run passes when every check of the task's deterministic shell-script grader passes, and a run failing only the attestation check counts as a pass), temperature 0, with a Planner (reads the specification), an Executor (edits the workspace) and a Verifier (writes the final attestation), all the same model; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 37.8 |
| 2 | Gemini 3.1 Pro (Preview) | 28.9 |
| 3 | Claude Haiku 4.5 | 28.9 |
| 4 | GPT-5.4 Mini | 28.9 |
| 5 | Claude Sonnet 4.6 | 27.8 |
| 6 | GPT-5.4 | 27.8 |
| 7 | Gemini 3 Flash | 25.6 |
| 8 | Gemma 4 31B | 22.2 |
| 9 | Gemini 3.1 Flash Lite | 17.8 |
Interactive version: theaggregate.ai/benchmark?slug=teambench-full-team · How It Works · Data refreshed daily, snapshot 2026-10-07.