Terminal-Bench 4.0

Artificial Analysis · 50 models

Rank	Model	Score
1	Kimi K3.1 Pro	0.6717
2	Gemini 4 Carbon (High)	0.6515
3	Claude Sonnet 5.5 (Max, Default Fallback)	0.6364
4	Gemini 4 Carbon (Low)	0.6263
5	Gemini 4 Carbon (Medium)	0.6061
6	Claude Opus 5.5 (Max, Default Fallback)	0.596
6	Claude Opus 5.5 (Xhigh, Default Fallback)	0.596
6	GPT-6 Astra (Xhigh)	0.596
9	GPT-6 Astra (Max)	0.5909
10	Claude Sonnet 5.5 (Xhigh, Default Fallback)	0.5707
10	Gemini 4 Argon (High)	0.5707
10	Muse Wonder Full (Max)	0.5707
13	Claude Opus 5.5 (High, Default Fallback)	0.5657
14	GPT-6.1 Sol (Max)	0.5606
15	GPT-6.1 Terra (Max)	0.5556
16	Claude Fable 5.1 (Xhigh, Default Fallback)	0.5505
17	GPT-6 Astra (High)	0.5404
17	GPT-6.1 Sol (Xhigh)	0.5404
19	Claude Opus 5.5 (Medium, Default Fallback)	0.5253
20	Claude Fable 5.1 (High, Default Fallback)	0.5202
20	Claude Fable 5.1 (Max, Default Fallback)	0.5202
20	Muse Wonder (Max)	0.5202
23	GPT-6.1 Sol (High)	0.5152
24	GPT-6 Astra (Medium)	0.4949
25	Claude Opus 5 (Max)	0.4899
26	GPT-6.1 Sol (Medium)	0.4798
27	Claude Opus 5 (Xhigh)	0.4646
27	GPT-6.1 Terra (Xhigh)	0.4646
29	Claude Opus 5 (High)	0.4596
30	Claude Fable 5.1 (Medium, Default Fallback)	0.4495
31	Claude Sonnet 5.5 (High, Default Fallback)	0.4394
31	GPT-6 Sol (Max)	0.4394
33	Claude Fable 5 (Max, Opus 4.8 Fallback)	0.4242
34	GLM-5.3 (Max)	0.4192
34	GPT-6 Astra (Low)	0.4192
34	GPT-6.1 Terra (High)	0.4192
37	Claude Fable 5.1 (Low, Default Fallback)	0.404
38	GPT-5.6 Sol (Max)	0.399
39	Qwen3.8 Max (0902)	0.3889
40	GPT-5.6 Terra (Max)	0.3535
41	GLM-5.3 (Low)	0.3485
41	MiMo-V2.6-Pro	0.3485
43	Claude Opus 5 (Medium)	0.3434
44	DeepSeek V4.1 Flash 1017 (Max)	0.3384
45	Ling 3.1 Flash	0.3333
45	Muse Spark 1.3 (Max)	0.3333
45	Step 5 Preview	0.3333
48	Claude Haiku 5.5 (Max)	0.3283
48	GLM-5.3-Flash	0.3283
48	Ling 3.1 Flash 1019	0.3283