Terminal-Bench 2.1 (Terminus-2)
Terminal-Bench 2.1 evaluated with Harbor/Terminus-2 framework, parser=json, temp=1.0, top_p=1.0, 128K context, 4h timeout, 32 CPU/48GB RAM, avg of 5 runs.
Category
Agentic Coding
Max Score
100.0
Score Type
percent
Active
Yes
Model Scores
Scores for Terminal-Bench 2.1 (Terminus-2)
| Model | Score | Date | Verified |
|---|---|---|---|
| Ornith-1.0-9B |
32.60%
|
25.06.2026 | Verified |
| Qwen3.5-9B |
0.44%
|
25.06.2026 | Unverified |
| Qwen3.5-35B-A3B |
30.09%
|
25.06.2026 | Unverified |
| Gemma4-12B |
21.00
|
25.06.2026 | Unverified |
| Gemma4-31B-it |
31.12%
|
25.06.2026 | Unverified |
| Ornith-1.0-35B |
63.72%
|
25.06.2026 | Verified |
| Qwen3.6-35B-A3B |
46.46%
|
25.06.2026 | Unverified |
| Qwen3.5-397B-A17B |
47.94%
|
25.06.2026 | Unverified |
| NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 |
52.21%
|
04.06.2026 | Verified |
| MiniMax-M2.7 |
50.89%
|
04.06.2026 | Unverified |
| GLM-5.1 |
56.49%
|
04.06.2026 | Unverified |
| Kimi K2.6 |
68.14%
|
04.06.2026 | Unverified |
| DeepSeek-V4-Pro |
41.59%
|
04.06.2026 | Unverified |
| DeepSeek-V4-Flash |
48.97%
|
04.06.2026 | Unverified |
| Qwen3.8-27B |
76.70%
|
18.08.2026 | Verified |
| Qwen3.6-27B |
62.54%
|
18.08.2026 | Unverified |
| Qwen3.7-Plus |
63.42%
|
18.08.2026 | Unverified |
| Muse Glimmer-30B |
45.28%
|
18.08.2026 | Unverified |
| Opus4.6 Max |
84.37%
|
18.08.2026 | Unverified |
| Muse-Glimmer-30B |
45.28%
|
01.08.2026 | Verified |
| GLM-5.2 |
88.50%
|
17.06.2026 | Verified |
| Gemini 3.1 Pro |
78.17%
|
17.06.2026 | Unverified |
| GPT-5.5 |
92.92%
|
17.06.2026 | Unverified |
| Claude Opus 4.8 |
94.40%
|
17.06.2026 | Unverified |
| Qwen3.7-Max |
79.65%
|
17.06.2026 | Unverified |
| MiniMax-M3 |
64.90%
|
17.06.2026 | Unverified |
| Ornith-1.5-35B-A3B |
69.03%
|
19.08.2026 | Verified |
| Ornith-1.0-35B-A3B |
63.72%
|
19.08.2026 | Unverified |
| Gemma-4-31B |
31.12%
|
19.08.2026 | Unverified |
| Hy4 preview |
94.99%
|
28.08.2026 | Verified |
| DeepSeek V4 Pro |
98.67%
|
28.08.2026 | Verified |
| Qwen3.8-2.4T-A95B |
96.76%
|
28.08.2026 | Verified |
| GLM-5.3 |
99.11%
|
28.08.2026 | Verified |
| Kimi K3 |
99.26%
|
28.08.2026 | Verified |
| GPT-5.6 Sol |
100.00%
|
28.08.2026 | Verified |
| Claude Opus 5 |
96.90%
|
28.08.2026 | Verified |
| Hy3 |
73.45%
|
28.08.2026 | Verified |