Terminal-Bench 2.1

Artificial Analysis · 50 models

Rank	Model	Score
1	Claude Fable 5.1 (Max, Default Fallback)	0.9139
2	Claude Fable 5.1 (Xhigh, Default Fallback)	0.9101
3	Claude Fable 5.1 (High, Default Fallback)	0.8989
3	GPT-6 Astra (High)	0.8989
5	GPT-5.6 Sol (Xhigh)	0.8951
5	GPT-6 Astra (Medium)	0.8951
7	Claude Opus 5 (Max)	0.8914
7	GPT-6 Astra (Xhigh)	0.8914
9	Qwen3.8 Max (0902)	0.8876
10	GPT-6 Astra (Max)	0.8839
10	Grok 4.6 (High)	0.8839
12	Claude Fable 5.1 (Medium, Default Fallback)	0.8801
12	Claude Opus 5 (Xhigh)	0.8801
12	GPT-5.6 Sol (Max)	0.8801
12	GPT-5.6 Terra (Max)	0.8801
12	GPT-6 Astra (Low)	0.8801
12	Grok 4.6 (Xhigh)	0.8801
18	Claude Opus 5 (High)	0.8764
18	Gemini 3.8 Flash (High)	0.8764
20	GPT-5.6 Sol (High)	0.8727
21	Claude Opus 5 (Medium)	0.8614
21	GPT-5.6 Sol (Medium)	0.8614
21	Qwen3.8-Flash-Next	0.8614
24	Gemini 3.7 Flash (High)	0.8577
25	Muse Spark 1.3 (Xhigh)	0.8539
26	Claude Fable 5.1 (Low, Default Fallback)	0.8502
26	Kimi K3 (Max)	0.8502
28	Claude Fable 5 (Max, Opus 4.8 Fallback)	0.8464
28	Claude Opus 4.8 (Max)	0.8464
30	GLM-5.3-Flash	0.8427
30	GPT-5.5 (Xhigh)	0.8427
30	Grok 4.6 (Medium)	0.8427
30	Muse Spark 1.3 (Max)	0.8427
34	GLM-5.3 (Max)	0.839
34	Gemini 3.8 Flash (Medium)	0.839
36	Claude Opus 4.7 (Max)	0.8315
36	Gemini 3.8 Flash (Low)	0.8315
38	Kimi K3 (Low)	0.824
39	Qwen3.8 2.4T A95B	0.8202
40	Grok 4.5 (High)	0.8165
41	Qwen3.8 Max	0.8127
42	GPT-5.6 Luna (Max)	0.809
43	Claude Sonnet 5 (Max)	0.8052
43	GPT-5.5 (Medium)	0.8052
45	GPT-5.6 Terra (Xhigh)	0.8015
45	Muse Spark 1.2 (Xhigh)	0.8015
47	Gemini 3.7 Flash (Low)	0.7978
47	Qwen3.8 27B (Xhigh)	0.7978
49	GPT-5.5 (High)	0.794
50	DeepSeek V4 Flash 0731 (Max)	0.7865