Terminal Bench 2.1
Terminal Bench 2.1 - agentic coding and terminal use benchmark. Evaluated with Claude Code (avg@10), 5-hour timeout, max_tokens=131,072.
Category
Agentic Coding
Max Score
—
Score Type
percent
Active
Yes
Model Scores
Scores for Terminal Bench 2.1
| Model | Score | Date | Verified |
|---|---|---|---|
| GLM 5.1 Thinking |
53.10%
|
11.06.2026 | Unverified |
| GPT 5.5 |
86.25%
|
11.06.2026 | Unverified |
| Kimi K2.6 |
59.31%
|
11.06.2026 | Unverified |
| MiniMax-M2.7 |
56.21%
|
11.06.2026 | Unverified |
| Qwen3.8-2.4T-A95B |
95.57%
|
18.08.2026 | Verified |
| Fable-5 |
93.35%
|
18.08.2026 | Unverified |
| GPT-5.6 Sol |
98.00%
|
18.08.2026 | Unverified |
| Claude Opus 4.8 |
93.35%
|
18.08.2026 | Unverified |
| Qwen3.7-Max |
82.15%
|
18.08.2026 | Unverified |
| DeepSeek V4 Pro |
65.63%
|
11.06.2026 | Unverified |
| Claude Opus 4.7 |
72.84%
|
11.06.2026 | Unverified |
| MiniMax-M3 |
72.73%
|
11.06.2026 | Unverified |
| Gemini 3.1 Pro |
77.49%
|
11.06.2026 | Unverified |
| Kimi K3 |
97.45%
|
18.08.2026 | Verified |
| DeepSeek-V4-Flash-0731 |
91.24%
|
31.07.2026 | Verified |
| DeepSeek-V4-Flash |
68.07%
|
13.08.2026 | Verified |
| GLM-5.2 |
89.36%
|
31.07.2026 | Verified |
| Opus-4.8 |
93.79%
|
31.07.2026 | Verified |
| DeepSeek-V4-Pro |
79.49%
|
13.08.2026 | Verified |
| DeepSeek-V4-Pro-0813 |
97.01%
|
13.08.2026 | Verified |
| GLM-5.3-Flash |
93.02%
|
27.08.2026 | Verified |
| DeepSeek-V4-Vision-Exp |
92.57%
|
27.08.2026 | Verified |
| GPT-5.3 Terra |
96.45%
|
27.08.2026 | Verified |
| Gemini 3.7 Flash |
94.68%
|
27.08.2026 | Verified |
| GLM-5.3 |
97.34%
|
28.08.2026 | Verified |
| DeepSeek-V4-Flash-Vision-Exp |
92.57%
|
08.09.2026 | Verified |
| LFM2.5-2.6B |
4.55%
|
— | Verified |
| Qwen3.5-2B |
2.88%
|
— | Verified |
| Qwen3.5-4B |
28.16%
|
— | Verified |
| granite-4.2-3B |
14.97%
|
— | Verified |
| Nemotron-3-Nano-4B |
3.77%
|
— | Verified |
| Gemma4-E4B |
1.66%
|
— | Verified |
| LFM2.5-8B-A1B |
1.66%
|
— | Verified |
| DeepSeek-V4.1-Flash |
100.00%
|
10.09.2026 | Verified |
| MiniCPM5-2B |
9.09%
|
— | Verified |
| Gemma4-E2B |
0.40
|
— | Verified |