Agents · the terminal · reference
Terminal-Bench 3.0
The 3.0 Terminal-Bench set: shell tasks an agent has to finish end to end. Scores can’t be compared across versions.
- Models
- 3
- By makers
- 3
- Latest
- Sep 10
| # | Model | Bar | Score | Source |
|---|---|---|---|---|
| 1 | DeepSeek V4.1 FlashDeepSeek · $0.30/1M · Coming soon | 30%vSep 10, 2026 | MakerSep 10, 2026 | |
| 2 | GLM 5.3Z.ai · $1.68/1M · Coming soon | 28.3%vAug 25, 2026 | MakerAug 25, 2026 | |
| 3 | Gemini 3.7 FlashGoogle · $1.50/1M · Coming soon | 14.9%vAug 13, 2026 | MakerAug 13, 2026 |
% of tasks passed. v marks a score the model’s maker reported. Prices are FullsendAI’s per million tokens, blended 3 input : 1 output.
About the data
From Terminal-Bench. Code and tasks are Apache-2.0; tbench.ai publishes no licence for its leaderboard data.