Benchmarks / Artificial Analysis

Reported by Artificial Analysis

Artificial Analysis

Share of terminal tasks (software, system and data jobs) an agent completes, as run by Artificial Analysis with its own harness and prompts.

Last updated 9 Nov 2026

Results dated
20 Oct 2026 to 9 Nov 2026
Results
204 configurations of 120 models
Unit
% of tasks
Licence
Artificial Analysis commercial data licence

Artificial Analysis in brief

What it measures. Artificial Analysis is a third-party benchmark, shown here with credit. Measured here: Share of terminal tasks (software, system and data jobs) an agent completes, as run by Artificial Analysis with its own harness and prompts.

Latest results. , 204 configurations ranked.

Top 3 on Terminal-Bench 4.0 (% of tasks, higher is better):

  1. Kimi K3.1 Pro (Moonshot AI): 67.2%
  2. Claude Sonnet 5.5 (max reasoning) (Anthropic): 63.6%
  3. Claude Opus 5.5 (max reasoning) (Anthropic): 59.6%

How to cite. Artificial Analysis, via Spring Prompt, results dated 9 November 2026. https://springprompt.com/benchmarks/artificial-analysis These results belong to Artificial Analysis, not to us. Licence: Artificial Analysis commercial data licence. Credit Artificial Analysis first.

Terminal-Bench 4.0: GPT-5.5

Top 15 of 204 results · % of tasks, higher is better. Choose a model to highlight it.Clear highlight

  1. 1 Kimi K3.1 ProMoonshot AI 67.2%
  2. 2 Claude Sonnet 5.5 (max reasoning)Anthropic 63.6%
  3. 3 Claude Opus 5.5 (max reasoning)Anthropic 59.6%
  4. 3 Claude Opus 5.5 (extra-high reasoning)Anthropic 59.6%
  5. 3 GPT-6 Astra (extra-high reasoning)OpenAI 59.6%
  6. 6 GPT-6 Astra (max reasoning)OpenAI 59.1%
  7. 7 Claude Sonnet 5.5 (extra-high reasoning)Anthropic 57.1%
  8. 7 Gemini 4 Argon (high reasoning)Google 57.1%
  9. 9 Claude Opus 5.5 (high reasoning)Anthropic 56.6%
  10. 10 GPT-6.1 Sol (max reasoning)OpenAI 56.1%
  11. 11 GPT-6.1 Terra (max reasoning)OpenAI 55.6%
  12. 12 Claude Fable 5.1 (extra-high reasoning)Anthropic 55.1%
  13. 13 GPT-6.1 Sol (extra-high reasoning)OpenAI 54.0%
  14. 13 GPT-6 Astra (high reasoning)OpenAI 54.0%
  15. 15 Claude Opus 5.5 (medium reasoning)Anthropic 52.5%
  16. 78 GPT-5.5 (extra-high reasoning)OpenAI 14.6%
  17. 98 GPT-5.5 (high reasoning)OpenAI 9.1%
  18. 110 GPT-5.5 (medium reasoning)OpenAI 5.1%

Full results

Artificial Analysis: Terminal-Bench 4.0, % of tasks, higher is better
#ModelTerminal-Bench 4.0
% of tasks, higher is better
Price
$ per million tokens, in / out
1 Kimi K3.1 ProMoonshot AI
67.2%
$1.75 / $10
2 Claude Sonnet 5.5 (max reasoning)Anthropic
63.6%
$2 / $10
3 Claude Opus 5.5 (max reasoning)Anthropic
59.6%
$4 / $20
3 Claude Opus 5.5 (extra-high reasoning)Anthropic
59.6%
$4 / $20
3 GPT-6 Astra (extra-high reasoning)OpenAI
59.6%
$10 / $50
6 GPT-6 Astra (max reasoning)OpenAI
59.1%
$10 / $50
7 Claude Sonnet 5.5 (extra-high reasoning)Anthropic
57.1%
$2 / $10
7 Gemini 4 Argon (high reasoning)Google
57.1%
$2 / $10
9 Claude Opus 5.5 (high reasoning)Anthropic
56.6%
$4 / $20
10 GPT-6.1 Sol (max reasoning)OpenAI
56.1%
$2 / $10
11 GPT-6.1 Terra (max reasoning)OpenAI
55.6%
$1 / $5
12 Claude Fable 5.1 (extra-high reasoning)Anthropic
55.1%
$10 / $50
13 GPT-6.1 Sol (extra-high reasoning)OpenAI
54.0%
$2 / $10
13 GPT-6 Astra (high reasoning)OpenAI
54.0%
$10 / $50
15 Claude Opus 5.5 (medium reasoning)Anthropic
52.5%
$4 / $20
16 Claude Fable 5.1 (high reasoning)Anthropic
52.0%
$10 / $50
16 Claude Fable 5.1 (max reasoning)Anthropic
52.0%
$10 / $50
16 Muse Wonder (max reasoning)Meta
52.0%
$4 / $20
19 GPT-6.1 Sol (high reasoning)OpenAI
51.5%
$2 / $10
20 GPT-6 Astra (medium reasoning)OpenAI
49.5%
$10 / $50
21 Claude Opus 5 (max reasoning)Anthropic
49.0%
$5 / $25
22 GPT-6.1 Sol (medium reasoning)OpenAI
48.0%
$2 / $10
23 GPT-6.1 Terra (extra-high reasoning)OpenAI
46.5%
$1 / $5
23 Claude Opus 5 (extra-high reasoning)Anthropic
46.5%
$5 / $25
25 Claude Opus 5 (high reasoning)Anthropic
46.0%
$5 / $25
26 Claude Fable 5.1 (medium reasoning)Anthropic
44.9%
$10 / $50
27 Claude Sonnet 5.5 (high reasoning)Anthropic
43.9%
$2 / $10
27 GPT-6 Sol (max reasoning)OpenAI
43.9%
$2 / $10
29 Claude Fable 5 (max reasoning)Anthropic
42.4%
$10 / $50
30 GPT-6 Astra (low reasoning)OpenAI
41.9%
$10 / $50
30 GLM-5.3 (max reasoning)Z.ai
41.9%
$1.40 / $4.40
30 GPT-6.1 Terra (high reasoning)OpenAI
41.9%
$1 / $5
33 Claude Fable 5.1 (low reasoning)Anthropic
40.4%
$10 / $50
34 GPT-5.6 Sol (max reasoning)OpenAI
39.9%
$4 / $20
35 Qwen3.8-Max (0902)Alibaba
38.9%
$2 / $6
36 Muse Wonder (high reasoning)Meta
38.4%
$4 / $20
37 GPT-5.6 Terra (max reasoning)OpenAI
35.4%
$2 / $12
38 MiMo-V2.6-ProXiaomi
34.8%
$0.43 / $0.87
38 GLM-5.3 (low reasoning)Z.ai
34.8%
$1.40 / $4.40
40 Claude Opus 5 (medium reasoning)Anthropic
34.3%
$5 / $25
41 DeepSeek-V4.1-Flash-1017 (max reasoning)DeepSeek
33.8%
$0.30 / $1.20
42 Muse Spark 1.3 (max reasoning)Meta
33.3%
$1.25 / $4.25
43 Claude Haiku 5.5 (max reasoning)Anthropic
32.8%
$0.10 / $0.50
43 GLM-5.3-FlashZ.ai
32.8%
$0.15 / $0.50
45 GPT-6.1 Terra (medium reasoning)OpenAI
32.3%
$1 / $5
46 Claude Opus 5.5 (low reasoning)Anthropic
31.3%
$4 / $20
47 GPT-6.1 Sol (low reasoning)OpenAI
30.8%
$2 / $10
48 GPT-6 Sol (extra-high reasoning)OpenAI
30.3%
$2 / $10
49 Claude Sonnet 5.5 (medium reasoning)Anthropic
29.8%
$2 / $10
50 Claude Haiku 5.5 (extra-high reasoning)Anthropic
29.3%
$0.10 / $0.50
51 DeepSeek-V4.1-Flash-1017 (high reasoning)DeepSeek
28.8%
$0.30 / $1.20
52 GPT-6.1 Terra (no reasoning)OpenAI
26.8%
$1 / $5
52 DeepSeek-V4.1-Flash (max reasoning)DeepSeek
26.8%
$0.30 / $1.20
54 Claude Opus 5 (low reasoning)Anthropic
26.3%
$5 / $25
54 GPT-6 Sol (high reasoning)OpenAI
26.3%
$2 / $10
56 MiniMax M3.1 FlashMiniMax
25.8%
$0.15 / $0.60
56 Grok 4.7 (extra-high reasoning)xAI
25.8%
$2 / $6
58 Qwen3.8-Flash-NextAlibaba
25.3%
–
59 GPT-5.6 Sol (extra-high reasoning)OpenAI
24.7%
$4 / $20
59 Grok 4.7 (high reasoning)xAI
24.7%
$2 / $6
59 Mistral Large 4Mistral AI
24.7%
$1.36 / $4.18
62 Mistral Large 4 (medium reasoning)Mistral AI
23.2%
$1.36 / $4.18
63 MiMo-V2.6-FlashXiaomi
22.7%
$0.14 / $0.28
64 Claude Haiku 5.5 (high reasoning)Anthropic
21.7%
$0.10 / $0.50
64 Claude Opus 4.8 (max reasoning)Anthropic
21.7%
$5 / $25
66 Grok 4.6 (high reasoning)xAI
21.2%
$2 / $6
67 Claude Sonnet 5.5 (low reasoning)Anthropic
20.7%
$2 / $10
67 GPT-5.6 Sol (high reasoning)OpenAI
20.7%
$4 / $20
69 Gemini 3.8 Flash (high reasoning)Google
19.7%
$1.50 / $7.50
69 Gemini 3.8 Flash (medium reasoning)Google
19.7%
$1.50 / $7.50
71 Qwen3.8-Max (0803)Alibaba
18.7%
–
71 GPT-6 Sol (medium reasoning)OpenAI
18.7%
$2 / $10
73 Grok 4.6 (extra-high reasoning)xAI
17.2%
$2 / $6
74 Muse Spark 1.3 (extra-high reasoning)Meta
16.7%
$1.25 / $4.25
75 Grok 4.7 (low reasoning)xAI
16.2%
$2 / $6
76 GPT-6.1 Terra (low reasoning)OpenAI
15.2%
$1 / $5
76 Claude Haiku 5.5 (medium reasoning)Anthropic
15.2%
$0.10 / $0.50
78 DeepSeek-V4-Pro (0423, max reasoning)DeepSeek
14.6%
$1.42 / $2.83
78 GPT-5.5 (extra-high reasoning)OpenAI
14.6%
$5 / $30
78 GPT-5.6 Sol (medium reasoning)OpenAI
14.6%
$4 / $20
81 Claude Sonnet 5 (max reasoning)Anthropic
14.1%
$2 / $10
81 DeepSeek-V4-Pro (0813, max reasoning)DeepSeek
14.1%
$1.32 / $3.96
83 Gemini 3.7 Flash (high reasoning)Google
13.6%
$1.50 / $7.50
84 GPT-6 Sol (no reasoning)OpenAI
13.1%
$2 / $10
84 Grok 4.6 (medium reasoning)xAI
13.1%
$2 / $6
86 Claude Haiku 5.5 (low reasoning)Anthropic
12.6%
$0.10 / $0.50
86 Kimi K3 (low reasoning)Moonshot AI
12.6%
$3 / $15
86 Kimi K3 (max reasoning)Moonshot AI
12.6%
$3 / $15
86 GPT-5.5 Instant (2026-06-26)OpenAI
12.6%
–
86 GPT-6 Luna (max reasoning)OpenAI
12.6%
$0.10 / $0.50
91 DeepSeek-V4-Flash (0731, max reasoning)DeepSeek
12.1%
$0.13 / $0.40
91 DeepSeek-V4-Flash-Vision-Exp (max reasoning)DeepSeek
12.1%
$0.22 / $0.65
93 GPT-5.6 Luna (max reasoning)OpenAI
11.6%
$0.20 / $1.20
94 Qwen3.8-2.4T-A95BAlibaba
11.1%
$2 / $6
95 Grok 4.5 (high reasoning)xAI
10.6%
$2 / $6
96 Gemini 3.8 Flash (low reasoning)Google
10.1%
$1.50 / $7.50
96 GPT-5.6 Terra (extra-high reasoning)OpenAI
10.1%
$2 / $12
98 DeepSeek-V4.1-Flash-1017 (no reasoning)DeepSeek
9.1%
$0.30 / $1.20
98 GPT-5.5 (high reasoning)OpenAI
9.1%
$5 / $30
98 GPT-6 Sol (low reasoning)OpenAI
9.1%
$2 / $10
101 Muse Spark Open (extra-high reasoning)Meta
8.1%
$1.25 / $4.25
101 GPT-6 Luna (extra-high reasoning)OpenAI
8.1%
$0.10 / $0.50
103 Claude Sonnet 5 (extra-high reasoning)Anthropic
7.1%
$2 / $10
103 Gemini 3.6 Flash (high reasoning)Google
7.1%
$1.50 / $7.50
103 Muse Spark 1.2 (extra-high reasoning)Meta
7.1%
$1.25 / $4.25
106 Gemini 3.5 Flash (high reasoning)Google
6.6%
$1.50 / $9
107 Muse Spark 1.1 (extra-high reasoning)Meta
6.1%
$1.25 / $4.25
108 Qwen3.8-27B (extra-high reasoning)Alibaba
5.6%
$0.50 / $3
108 DeepSeek-V4.1-Flash (no reasoning)DeepSeek
5.6%
$0.30 / $1.20
110 Qwen3.8-27B (medium reasoning)Alibaba
5.1%
$0.50 / $3
110 Claude Sonnet 5 (high reasoning)Anthropic
5.1%
$2 / $10
110 GPT-5.5 (medium reasoning)OpenAI
5.1%
$5 / $30
113 GPT-6 Luna (high reasoning)OpenAI
4.5%
$0.10 / $0.50
114 Gemini 3.1 Pro PreviewGoogle
4.0%
$2 / $12
115 GPT-5.6 Luna (extra-high reasoning)OpenAI
3.5%
$0.20 / $1.20
116 Claude Sonnet 4.6 (max reasoning)Anthropic
3.0%
$3 / $15
116 DeepSeek-V4-Flash (0423, high reasoning)DeepSeek
3.0%
$0.05 / $0.45
116 Beam 1.0 501B A23BReflection
3.0%
$0.40 / $1.20
116 Grok 4.6 (low reasoning)xAI
3.0%
$2 / $6
120 Qwen3.8-27B (low reasoning)Alibaba
2.5%
$0.50 / $3
120 Claude Sonnet 5 (low reasoning)Anthropic
2.5%
$2 / $10
120 DeepSeek-V4-Flash (0423, max reasoning)DeepSeek
2.5%
$0.05 / $0.45
120 GPT-5.6 Luna (high reasoning)OpenAI
2.5%
$0.20 / $1.20
120 GPT-6 Luna (medium reasoning)OpenAI
2.5%
$0.10 / $0.50
125 Claude Sonnet 5 (medium reasoning)Anthropic
2.0%
$2 / $10
125 MiniMax-M3MiniMax
2.0%
$0.30 / $1.20
125 GPT-5.4 mini (extra-high reasoning)OpenAI
2.0%
$0.75 / $4.50
125 GLM-5.1Z.ai
2.0%
$1.33 / $4.18
129 Qwen3.7-MaxAlibaba
1.5%
$1.48 / $4.42
129 DeepSeek-V4-Pro (0813, no reasoning)DeepSeek
1.5%
$1.32 / $3.96
129 GPT-5.6 Terra (high reasoning)OpenAI
1.5%
$2 / $12
129 GPT-5.6 Terra (low reasoning)OpenAI
1.5%
$2 / $12
129 GPT-6 Luna (no reasoning)OpenAI
1.5%
$0.10 / $0.50
134 Qwen3.7-PlusAlibaba
1.0%
$0.32 / $1.28
134 Gemini 3.5 Flash-LiteGoogle
1.0%
$0.30 / $2.50
134 Kimi K2.7 CodeMoonshot AI
1.0%
$0.95 / $4
134 GPT-5.6 Luna (no reasoning)OpenAI
1.0%
$0.20 / $1.20
134 GPT-5.6 Sol (low reasoning)OpenAI
1.0%
$4 / $20
134 GPT-5.6 Terra (medium reasoning)OpenAI
1.0%
$2 / $12
134 Inkling SmallThinking Machines
1.0%
$0.45 / $1.20
134 Inkling (extra-high reasoning)Thinking Machines
1.0%
$1 / $4.05
134 GLM-5.2 (max reasoning)Z.ai
1.0%
$1.40 / $4.40
143 Qwen3.5-9BAlibaba
0.5%
$0.10 / $0.15
143 Trinity Large ThinkingArcee AI
0.5%
–
143 Gemini 3.1 Flash-Lite PreviewGoogle
0.5%
$0.25 / $1.50
143 Muse Glimmer (high reasoning)Meta
0.5%
–
143 Kimi K2.6Moonshot AI
0.5%
$0.95 / $4
143 GPT-5.4 nano (extra-high reasoning)OpenAI
0.5%
$0.20 / $1.25
143 GPT-5.6 Luna (medium reasoning)OpenAI
0.5%
$0.20 / $1.20
143 GPT-5.6 Terra (no reasoning)OpenAI
0.5%
$2 / $12
143 Hy3Tencent
0.5%
$0.14 / $0.58
152 Qwen3-14B (reasoning on)Alibaba
0.0%
$0.10 / $0.22
152 Qwen3-235B-A22B-Thinking-2507Alibaba
0.0%
$0.45 / $3.50
152 Qwen3-30B-A3B-Thinking-2507Alibaba
0.0%
–
152 Qwen3-32B (reasoning on)Alibaba
0.0%
$0.14 / $0.40
152 Qwen3.5-122B-A10BAlibaba
0.0%
$0.26 / $2.08
152 Qwen3.5-397B-A17BAlibaba
0.0%
$0.55 / $3.50
152 Qwen3.6-27BAlibaba
0.0%
$0.30 / $3.20
152 Qwen3.6-35B-A3BAlibaba
0.0%
$0.10 / $0.95
152 Qwen3.8-27B (no reasoning)Alibaba
0.0%
$0.50 / $3
152 Qwen3-8B (reasoning on)Alibaba
0.0%
–
152 Qwen3-Coder-NextAlibaba
0.0%
$0.12 / $0.80
152 Claude Haiku 4.5 (reasoning on)Anthropic
0.0%
$1 / $5
152 Claude Sonnet 4.5 (reasoning on)Anthropic
0.0%
$3 / $15
152 DeepSeek-V3-0324DeepSeek
0.0%
$0.25 / $1
152 DeepSeek-V3DeepSeek
0.0%
$0.26 / $1.03
152 DeepSeek-R1DeepSeek
0.0%
–
152 DeepSeek-V3.1-Terminus (reasoning on)DeepSeek
0.0%
$0.27 / $1
152 Gemini 2.5 ProGoogle
0.0%
$1.25 / $10
152 Gemma 3 12BGoogle
0.0%
$0.05 / $0.15
152 Gemma 3 27BGoogle
0.0%
$0.08 / $0.35
152 Gemma 4 31BGoogle
0.0%
$0.15 / $0.40
152 Gemma 4 E4BGoogle
0.0%
–
152 Granite 4.2 8BIBM
0.0%
$0.06 / $0.15
152 Mercury 2Inception
0.0%
–
152 Llama 4 MaverickMeta
0.0%
$0.35 / $1
152 Llama 4 ScoutMeta
0.0%
$0.10 / $0.30
152 MiniMax-M2.7MiniMax
0.0%
$0.30 / $1.20
152 Devstral 2Mistral AI
0.0%
$0.40 / $2
152 Devstral Small 2Mistral AI
0.0%
–
152 Ministral 3 14BMistral AI
0.0%
$0.20 / $0.20
152 Ministral 3 3BMistral AI
0.0%
$0.10 / $0.10
152 Ministral 3 8BMistral AI
0.0%
$0.15 / $0.15
152 Mistral Large 3Mistral AI
0.0%
$0.50 / $1.50
152 Mistral Medium 3.1Mistral AI
0.0%
$0.40 / $2
152 Mistral Medium 3.5Mistral AI
0.0%
$1.50 / $7.50
152 Mistral Small 4Mistral AI
0.0%
$0.15 / $0.60
152 Mistral Small 3.1 24BMistral AI
0.0%
$0.35 / $0.56
152 Mistral Small 3.2 24BMistral AI
0.0%
$0.094 / $0.25
152 Nemotron 3.5 Super VLNVIDIA
0.0%
$0.15 / $0.50
152 Nemotron 3 Nano 30B A3B (reasoning on)NVIDIA
0.0%
$0.06 / $0.24
152 Nemotron 3 Super 120B A12BNVIDIA
0.0%
$0.08 / $0.45
152 GPT-5.6 Luna (low reasoning)OpenAI
0.0%
$0.20 / $1.20
152 GPT-5 mini (high reasoning)OpenAI
0.0%
$0.25 / $2
152 GPT-6 Luna (low reasoning)OpenAI
0.0%
$0.10 / $0.50
152 gpt-oss-120b (high reasoning)OpenAI
0.0%
$0.10 / $0.50
152 gpt-oss-20b (high reasoning)OpenAI
0.0%
$0.03 / $0.14
152 o3-mini (high reasoning)OpenAI
0.0%
–
152 Solar Pro 3Upstage
0.0%
–
152 Solar Pro 4Upstage
0.0%
$0.15 / $0.60
152 Grok 4.3 (high reasoning)xAI
0.0%
$1.25 / $2.50
152 Grok 4.3 (no reasoning)xAI
0.0%
$1.25 / $2.50
152 MiMo-V2.5-Pro (reasoning on)Xiaomi
0.0%
$0.43 / $0.87
152 MiMo-V2.5Xiaomi
0.0%
$0.40 / $2

Swipe the table sideways for more columns.

Source: Artificial Analysis, via Spring Prompt, springprompt.com/benchmarks/artificial-analysis. Test your own prompt on these models →

Ranks follow the score as shown, so equal numbers share a rank. Results as published by Artificial Analysis; we do not re-run them.

Progress over time

Terminal-Bench 4.0, plotted at each model's release date, higher is better. Each model counts once, at its best setting. Scores are today's, not launch day's. Hover or tap a mark for its model.

Best score so far

Every model at its release date. The line moves each time a new release beats every earlier one, and each new best is named where there is room.

-20%0%20%40%60%80% 2025Jul2026Jul DeepSeek-V3Claude Sonnet 4.6Gemini 3.1 Pro PreviewGPT-5.5Claude Opus 4.8Claude Fable 5Claude Opus 5Claude Fable 5.1GPT-6 AstraClaude Sonnet 5.5Kimi K3.1 Pro

Best score by lab

Each lab's best so far, stepping whenever one of its releases beats its own earlier models. The 6 labs with the best scores today; 14 others not drawn.

Moonshot AIAnthropicOpenAIGoogleMetaZ.ai

-20%0%20%40%60%80% 2025Jul2026Jul Moonshot AIAnthropicOpenAIGoogleMetaZ.ai

What it measures

Share of terminal tasks (software, system and data jobs) an agent completes, as run by Artificial Analysis with its own harness and prompts.

What it does not measure

Not other harnesses or tools; one attempt per task.

291 results from Artificial Analysis not ranked here · show why

We rank a result only when we can tie it to a specific model you can use. These are left out:

  • Not on sale through the API providers we track: 276
  • A different snapshot or variant from the model we list: 14
  • An unusual combination of settings: 1

Data sourced from Artificial Analysis. Licence: Artificial Analysis commercial data licence.