artificialanalysis.ai · benchmark source

Artificial Analysis

Independent evaluations across composite intelligence, science reasoning, and agentic terminal tasks. Multiple boards live under one publisher.

Open artificialanalysis.ai3 boards on this publisher

Tool use

Terminal-Bench v2.1

A terminal-oriented evaluation of agents that must use shell tools and complete multi-step tasks in an environment.

Why it matters

It tests the loop around a model: planning, commands, file changes, and validation.

Use it when

You are choosing a terminal agent and want an independent harness signal about multi-step tool interaction.

The limitation

A high score does not guarantee safe behaviour in your repository. Local rows here use Artificial Analysis Terminus 2; proprietary agent harnesses on tbench.ai are separate.

Leaderboard snapshot

Terminal-Bench v2.1 results in a readable view

Use the tabs for the primary score and other published fields from this board. The source link remains authoritative for the live table.

Full Artificial Analysis Terminus 2 table mirrored here. Proprietary agent harnesses on tbench.ai are not mixed into this board.

Local leader

GPT-5.6 Sol (xhigh)

89.5%

Rows shown

191

Success rate

Snapshot date

2026-08-14

1 day old · not a live API feed

Score profile

Success rate · Artificial Analysis Terminus 2

Higher is better in this view

GPT-5.6 Sol (xhigh)

89.5%

Claude Opus 5 (Adaptive Reasoning, Max Effort)

89.1%

Grok 4.6 (high)

88.4%

Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)

88%

GPT-5.6 Sol (max)

88%

GPT-5.6 Terra (max)

88%

Claude Opus 5 (Adaptive Reasoning, High Effort)

87.6%

GPT-5.6 Sol (high)

87.3%

Claude Opus 5 (Adaptive Reasoning, Medium Effort)

86.1%

GPT-5.6 Sol (medium)

86.1%

Gemini 3.7 Flash (high)

85.8%

Kimi K3 (max)

85%

Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)

84.6%

Claude Opus 4.8 (Adaptive Reasoning, Max Effort)

84.6%

GPT-5.5 (xhigh)

84.3%

Claude Opus 4.7 (Adaptive Reasoning, Max Effort)

83.1%

Kimi K3 (low)

82.4%

Grok 4.5 (high)

81.6%

Qwen3.8 Max

81.3%

GPT-5.6 Luna (max)

80.9%

Claude Sonnet 5 (Adaptive Reasoning, Max Effort)

80.5%

GPT-5.5 (medium)

80.5%

GPT-5.6 Terra (xhigh)

80.1%

Muse Spark 1.2 (xhigh)

80.1%

GPT-5.5 (high)

79.4%

DeepSeek V4 Flash 0731 (Reasoning, Max Effort)

78.7%

DeepSeek V4 Pro 0813 (max)

78.7%

Gemini 3.5 Flash (high)

78.7%

GPT-5.4 (xhigh)

78.3%

GLM-5.2 (max)

77.9%

GPT-5.6 Luna (xhigh)

77.9%

Muse Spark 1.1 (xhigh)

77.9%

Gemini 3.6 Flash (high)

77.5%

GPT-5.6 Sol (low)

76.8%

Claude Opus 5 (Adaptive Reasoning, Low Effort)

76.4%

GPT-5.6 Terra (high)

75.7%

Claude Sonnet 5 (Non-reasoning, High Effort)

75.3%

Qwen3.7 Max

74.5%

GPT-5.6 Sol (Non-reasoning)

74.2%

Gemini 3.1 Pro Preview

73.8%

GPT-5.6 Terra (medium)

72.3%

Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

71.2%

Motif 3 (Beta)

70.8%

KAT Coder Pro V2

70%

GPT-5.6 Luna (high)

69.7%

Nex-N2-Pro

67.8%

Kimi K2.7 Code

67.4%

Agnes 2.5 Pro Alpha

67%

Kimi K2.6

65.9%

GPT-5.5 (low)

65.5%

MiMo-V2.5-Pro

65.2%

MiniMax-M3

65.2%

DeepSeek V4 Pro (Reasoning, High Effort)

64.8%

Hy3

64.4%

MiMo-V2.5

63.7%

GPT-5.6 Terra (low)

62.5%

Muse Spark

62.2%

DeepSeek V4 Flash (Reasoning, Max Effort)

61.8%

GLM-5.1 (Reasoning)

61.8%

MiMo-V2-Flash (Non-reasoning)

61.8%

Qwen3.6 Plus

61.4%

GPT-5.5 (Non-reasoning)

61%

Qwen3.7 Plus

61%

GPT-5.4 nano (xhigh)

60.7%

Qwen3.6 27B (Reasoning)

60.7%

JT-4.1 Flash 236B A21B

59.6%

GPT-5.4 mini (xhigh)

59.2%

DeepSeek V4 Flash (Reasoning, High Effort)

56.9%

GPT-5.6 Terra (Non-reasoning)

56.2%

Claude 4.5 Sonnet (Reasoning)

55.8%

Ling 3.0 Flash

55.4%

MiniMax-M2.7

55.4%

Inkling (xhigh)

55.1%

Inkling Small

55.1%

Nemotron 3 Ultra 550B A55B (Reasoning)

53.9%

Gemini 3.5 Flash-Lite

53.6%

GPT-5.6 Luna (medium)

53.2%

GPT-5.1 (high)

52.4%

Grok Build 0.1 0616

52.1%

GLM-5.2 (Non-reasoning)

51.7%

Muse Glimmer (high)

51.7%

Qwen3.5 397B A17B (Reasoning)

51.3%

Qwen3.6 27B (Non-reasoning)

51.3%

Mistral Medium 3.5

50.6%

LongCat 2.0

50.2%

GLM-4.6 (Reasoning)

49.4%

Qwen3.5 122B A10B (Reasoning)

47.6%

Qwen3.5 122B A10B (Non-reasoning)

47.2%

DeepSeek V3.2 (Reasoning)

46.8%

Kimi K2.5 (Reasoning)

45.7%

GLM-4.7 (Reasoning)

45.3%

DeepSeek V3.1 Terminus (Reasoning)

44.9%

Qwen3.6 35B A3B (Reasoning)

44.9%

Claude 4.5 Haiku (Reasoning)

44.2%

Gemma 4 31B (Reasoning)

43.4%

GPT-5.6 Luna (low)

43.4%

Ring-2.6-1T

43.1%

Qwen3.6 35B A3B (Non-reasoning)

41.6%

Qwen3.5 35B A3B (Non-reasoning)

40.8%

Grok 4.3 (high)

39.7%

Step 3.7 Flash

39.3%

Gemma 4 26B A4B (Reasoning)

39%

GPT-5.6 Luna (Non-reasoning)

39%

Nemotron 3 Super 120B A12B (Reasoning)

38.6%

Qwen3 Coder Next

38.2%

Claude 4 Sonnet (Reasoning)

36.3%

North Mini Code

35.6%

GPT-5 (high)

35.2%

GPT-5.5 Instant (June 2026)

34.8%

Grok 4.3 (Non-reasoning)

34.1%

Gemini 3.1 Flash-Lite

31.1%

Devstral 2

30.3%

K-EXAONE (Reasoning)

30.3%

Devstral Small 2

29.6%

Nova 2.0 Pro Preview (medium)

29.6%

Gemma 4 31B (Non-reasoning)

29.2%

Qwen3.5 9B (Reasoning)

29.2%

G9v3-39A5B

28.5%

Gemini 2.5 Pro

28.5%

Ling 3.0 Tiny

27.7%

Gemma 4 12B (Reasoning)

27.3%

Mercury 2

27.3%

gpt-oss-120b (high)

26.2%

Mistral Small 3.1

26.2%

Qwen3.5 4B (Reasoning)

25.8%

Ling 2.6 Flash

24.3%

Nemotron 3.5 Lightning

24.3%

Command A+

22.8%

EXAONE 4.5 33B

21.3%

Qwen3.5 4B (Non-reasoning)

21.3%

Qwen3.5 9B (Non-reasoning)

21.3%

Mistral Small 4 (Reasoning)

21%

Nemotron Cascade 2 30B A3B

20.6%

Trinity Large Thinking

20.6%

Nova 2.0 Pro Preview (low)

19.5%

DeepSeek R1 (Jan '25)

19.1%

HyperNova 60B 2605

18.4%

Nova 2.0 Pro Preview (Non-reasoning)

17.2%

DeepSeek V3 (Dec '24)

16.9%

Nova 2.0 Lite (high)

16.1%

K2 Think V2

15%

DeepSeek V3 0324

13.9%

gpt-oss-120b (low)

13.9%

gpt-oss-20b (high)

13.9%

Mistral Medium 3.1

13.9%

DiffusionGemma 26B A4B

12.4%

Magistral Medium 1.2

12.4%

Mistral Large 3

12%

Qwen3 235B A22B 2507 (Reasoning)

12%

Solar Pro 3

12%

Celeris-1

11.2%

Claude 3.5 Haiku

10.1%

GPT-4.1 mini

10.1%

Ministral 3 14B

9.7%

Llama 4 Maverick

7.9%

Nemotron 3 Nano Omni 30B A3B Reasoning

6.7%

NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)

6.7%

Qwen3 Next 80B A3B (Reasoning)

6.7%

G9v3-3B

GPT-4o mini

5.6%

Mistral Small 3.2

5.6%

Qwen3 32B (Reasoning)

5.2%

Llama 3.3 Instruct 70B

4.9%

Qwen3 14B (Reasoning)

4.9%

Gemma 3 27B Instruct

4.5%

Magistral Small 1.2

4.5%

o3-mini (high)

4.5%

Ministral 3 8B

4.1%

GPT-4.1 nano

3.7%

GPT-5 mini (high)

3.7%

Llama 4 Scout

3.7%

NVIDIA Nemotron 3 Nano 4B

3.7%

Granite 4.1 8B

3.4%

Qwen3.5 2B (Reasoning)

Granite 4.1 30B

2.6%

Qwen3 8B (Reasoning)

2.2%

Gemma 4 E4B (Reasoning)

1.9%

Llama 3.1 Instruct 8B

1.5%

Qwen3 30B A3B 2507 (Reasoning)

1.5%

Granite 4.1 3B

1.1%

Nanbeige4.1-3B

1.1%

Gemma 3n E4B Instruct

0.7%

Gemma 3 4B Instruct

0.4%

Gemma 4 E2B (Reasoning)

0.4%

Phi-4 Mini Instruct

0.4%

Qwen3.5 0.8B (Non-reasoning)

0.4%

Gemma 3 12B Instruct

MiniCPM-V 4.6 1.3B

Ministral 3 3B

Qwen3.5 0.8B (Reasoning)

Qwen3.5 2B (Non-reasoning)

Benchmark leaderboard snapshot showing rank, score, published details, and notes
Rank	Model or system	Success rate	Other details	Note
#1	GPT-5.6 Sol (xhigh)OpenAIArtificial Analysis Terminus 2	89.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#2	Claude Opus 5 (Adaptive Reasoning, Max Effort)AnthropicArtificial Analysis Terminus 2	89.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#3	Grok 4.6 (high)SpaceXAIArtificial Analysis Terminus 2	88.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#4	Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)AnthropicArtificial Analysis Terminus 2	88%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#4	GPT-5.6 Sol (max)OpenAIArtificial Analysis Terminus 2	88%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#4	GPT-5.6 Terra (max)OpenAIArtificial Analysis Terminus 2	88%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#7	Claude Opus 5 (Adaptive Reasoning, High Effort)AnthropicArtificial Analysis Terminus 2	87.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#8	GPT-5.6 Sol (high)OpenAIArtificial Analysis Terminus 2	87.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#9	Claude Opus 5 (Adaptive Reasoning, Medium Effort)AnthropicArtificial Analysis Terminus 2	86.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#9	GPT-5.6 Sol (medium)OpenAIArtificial Analysis Terminus 2	86.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#11	Gemini 3.7 Flash (high)GoogleArtificial Analysis Terminus 2	85.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#12	Kimi K3 (max)Moonshot AIArtificial Analysis Terminus 2	85%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#13	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)AnthropicArtificial Analysis Terminus 2	84.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#13	Claude Opus 4.8 (Adaptive Reasoning, Max Effort)AnthropicArtificial Analysis Terminus 2	84.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#15	GPT-5.5 (xhigh)OpenAIArtificial Analysis Terminus 2	84.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#16	Claude Opus 4.7 (Adaptive Reasoning, Max Effort)AnthropicArtificial Analysis Terminus 2	83.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#17	Kimi K3 (low)Moonshot AIArtificial Analysis Terminus 2	82.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#18	Grok 4.5 (high)SpaceXAIArtificial Analysis Terminus 2	81.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#19	Qwen3.8 MaxQwenArtificial Analysis Terminus 2	81.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#20	GPT-5.6 Luna (max)OpenAIArtificial Analysis Terminus 2	80.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#21	Claude Sonnet 5 (Adaptive Reasoning, Max Effort)AnthropicArtificial Analysis Terminus 2	80.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#21	GPT-5.5 (medium)OpenAIArtificial Analysis Terminus 2	80.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#23	GPT-5.6 Terra (xhigh)OpenAIArtificial Analysis Terminus 2	80.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#23	Muse Spark 1.2 (xhigh)MetaArtificial Analysis Terminus 2	80.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#25	GPT-5.5 (high)OpenAIArtificial Analysis Terminus 2	79.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#26	DeepSeek V4 Flash 0731 (Reasoning, Max Effort)DeepSeekArtificial Analysis Terminus 2	78.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#26	DeepSeek V4 Pro 0813 (max)DeepSeekArtificial Analysis Terminus 2	78.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#26	Gemini 3.5 Flash (high)GoogleArtificial Analysis Terminus 2	78.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#29	GPT-5.4 (xhigh)OpenAIArtificial Analysis Terminus 2	78.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#30	GLM-5.2 (max)Z.aiArtificial Analysis Terminus 2	77.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#30	GPT-5.6 Luna (xhigh)OpenAIArtificial Analysis Terminus 2	77.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#30	Muse Spark 1.1 (xhigh)MetaArtificial Analysis Terminus 2	77.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#33	Gemini 3.6 Flash (high)GoogleArtificial Analysis Terminus 2	77.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#34	GPT-5.6 Sol (low)OpenAIArtificial Analysis Terminus 2	76.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#35	Claude Opus 5 (Adaptive Reasoning, Low Effort)AnthropicArtificial Analysis Terminus 2	76.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#36	GPT-5.6 Terra (high)OpenAIArtificial Analysis Terminus 2	75.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#37	Claude Sonnet 5 (Non-reasoning, High Effort)AnthropicArtificial Analysis Terminus 2	75.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#38	Qwen3.7 MaxQwenArtificial Analysis Terminus 2	74.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#39	GPT-5.6 Sol (Non-reasoning)OpenAIArtificial Analysis Terminus 2	74.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#40	Gemini 3.1 Pro PreviewGoogleArtificial Analysis Terminus 2	73.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#41	GPT-5.6 Terra (medium)OpenAIArtificial Analysis Terminus 2	72.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#42	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)AnthropicArtificial Analysis Terminus 2	71.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#43	Motif 3 (Beta)UnknownArtificial Analysis Terminus 2	70.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#44	KAT Coder Pro V2UnknownArtificial Analysis Terminus 2	70%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#45	GPT-5.6 Luna (high)OpenAIArtificial Analysis Terminus 2	69.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#46	Nex-N2-ProUnknownArtificial Analysis Terminus 2	67.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#47	Kimi K2.7 CodeMoonshot AIArtificial Analysis Terminus 2	67.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#48	Agnes 2.5 Pro AlphaUnknownArtificial Analysis Terminus 2	67%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#49	Kimi K2.6Moonshot AIArtificial Analysis Terminus 2	65.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#50	GPT-5.5 (low)OpenAIArtificial Analysis Terminus 2	65.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#51	MiMo-V2.5-ProXiaomiArtificial Analysis Terminus 2	65.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#51	MiniMax-M3MiniMaxArtificial Analysis Terminus 2	65.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#53	DeepSeek V4 Pro (Reasoning, High Effort)DeepSeekArtificial Analysis Terminus 2	64.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#54	Hy3UnknownArtificial Analysis Terminus 2	64.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#55	MiMo-V2.5XiaomiArtificial Analysis Terminus 2	63.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#56	GPT-5.6 Terra (low)OpenAIArtificial Analysis Terminus 2	62.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#57	Muse SparkMetaArtificial Analysis Terminus 2	62.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#58	DeepSeek V4 Flash (Reasoning, Max Effort)DeepSeekArtificial Analysis Terminus 2	61.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#58	GLM-5.1 (Reasoning)Z.aiArtificial Analysis Terminus 2	61.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#58	MiMo-V2-Flash (Non-reasoning)XiaomiArtificial Analysis Terminus 2	61.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#61	Qwen3.6 PlusQwenArtificial Analysis Terminus 2	61.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#62	GPT-5.5 (Non-reasoning)OpenAIArtificial Analysis Terminus 2	61%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#62	Qwen3.7 PlusQwenArtificial Analysis Terminus 2	61%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#64	GPT-5.4 nano (xhigh)OpenAIArtificial Analysis Terminus 2	60.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#64	Qwen3.6 27B (Reasoning)QwenArtificial Analysis Terminus 2	60.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#66	JT-4.1 Flash 236B A21BUnknownArtificial Analysis Terminus 2	59.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#67	GPT-5.4 mini (xhigh)OpenAIArtificial Analysis Terminus 2	59.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#68	DeepSeek V4 Flash (Reasoning, High Effort)DeepSeekArtificial Analysis Terminus 2	56.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#69	GPT-5.6 Terra (Non-reasoning)OpenAIArtificial Analysis Terminus 2	56.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#70	Claude 4.5 Sonnet (Reasoning)AnthropicArtificial Analysis Terminus 2	55.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#71	Ling 3.0 FlashUnknownArtificial Analysis Terminus 2	55.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#71	MiniMax-M2.7MiniMaxArtificial Analysis Terminus 2	55.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#73	Inkling (xhigh)InclusionAIArtificial Analysis Terminus 2	55.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#73	Inkling SmallInclusionAIArtificial Analysis Terminus 2	55.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#75	Nemotron 3 Ultra 550B A55B (Reasoning)NVIDIAArtificial Analysis Terminus 2	53.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#76	Gemini 3.5 Flash-LiteGoogleArtificial Analysis Terminus 2	53.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#77	GPT-5.6 Luna (medium)OpenAIArtificial Analysis Terminus 2	53.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#78	GPT-5.1 (high)OpenAIArtificial Analysis Terminus 2	52.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#79	Grok Build 0.1 0616SpaceXAIArtificial Analysis Terminus 2	52.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#80	GLM-5.2 (Non-reasoning)Z.aiArtificial Analysis Terminus 2	51.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#80	Muse Glimmer (high)MetaArtificial Analysis Terminus 2	51.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#82	Qwen3.5 397B A17B (Reasoning)QwenArtificial Analysis Terminus 2	51.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#82	Qwen3.6 27B (Non-reasoning)QwenArtificial Analysis Terminus 2	51.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#84	Mistral Medium 3.5MistralArtificial Analysis Terminus 2	50.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#85	LongCat 2.0TencentArtificial Analysis Terminus 2	50.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#86	GLM-4.6 (Reasoning)Z.aiArtificial Analysis Terminus 2	49.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#87	Qwen3.5 122B A10B (Reasoning)QwenArtificial Analysis Terminus 2	47.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#88	Qwen3.5 122B A10B (Non-reasoning)QwenArtificial Analysis Terminus 2	47.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#89	DeepSeek V3.2 (Reasoning)DeepSeekArtificial Analysis Terminus 2	46.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#90	Kimi K2.5 (Reasoning)Moonshot AIArtificial Analysis Terminus 2	45.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#91	GLM-4.7 (Reasoning)Z.aiArtificial Analysis Terminus 2	45.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#92	DeepSeek V3.1 Terminus (Reasoning)DeepSeekArtificial Analysis Terminus 2	44.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#92	Qwen3.6 35B A3B (Reasoning)QwenArtificial Analysis Terminus 2	44.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#94	Claude 4.5 Haiku (Reasoning)AnthropicArtificial Analysis Terminus 2	44.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#95	Gemma 4 31B (Reasoning)GoogleArtificial Analysis Terminus 2	43.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#95	GPT-5.6 Luna (low)OpenAIArtificial Analysis Terminus 2	43.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#97	Ring-2.6-1TUnknownArtificial Analysis Terminus 2	43.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#98	Qwen3.6 35B A3B (Non-reasoning)QwenArtificial Analysis Terminus 2	41.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#99	Qwen3.5 35B A3B (Non-reasoning)QwenArtificial Analysis Terminus 2	40.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#100	Grok 4.3 (high)SpaceXAIArtificial Analysis Terminus 2	39.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#101	Step 3.7 FlashUnknownArtificial Analysis Terminus 2	39.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#102	Gemma 4 26B A4B (Reasoning)GoogleArtificial Analysis Terminus 2	39%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#102	GPT-5.6 Luna (Non-reasoning)OpenAIArtificial Analysis Terminus 2	39%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#104	Nemotron 3 Super 120B A12B (Reasoning)NVIDIAArtificial Analysis Terminus 2	38.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#105	Qwen3 Coder NextQwenArtificial Analysis Terminus 2	38.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#106	Claude 4 Sonnet (Reasoning)AnthropicArtificial Analysis Terminus 2	36.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#107	North Mini CodeUnknownArtificial Analysis Terminus 2	35.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#108	GPT-5 (high)OpenAIArtificial Analysis Terminus 2	35.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#109	GPT-5.5 Instant (June 2026)OpenAIArtificial Analysis Terminus 2	34.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#110	Grok 4.3 (Non-reasoning)SpaceXAIArtificial Analysis Terminus 2	34.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#111	Gemini 3.1 Flash-LiteGoogleArtificial Analysis Terminus 2	31.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#112	Devstral 2MistralArtificial Analysis Terminus 2	30.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#112	K-EXAONE (Reasoning)UpstageArtificial Analysis Terminus 2	30.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#114	Devstral Small 2MistralArtificial Analysis Terminus 2	29.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#114	Nova 2.0 Pro Preview (medium)AmazonArtificial Analysis Terminus 2	29.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#116	Gemma 4 31B (Non-reasoning)GoogleArtificial Analysis Terminus 2	29.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#116	Qwen3.5 9B (Reasoning)QwenArtificial Analysis Terminus 2	29.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#118	G9v3-39A5BUnknownArtificial Analysis Terminus 2	28.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#118	Gemini 2.5 ProGoogleArtificial Analysis Terminus 2	28.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#120	Ling 3.0 TinyUnknownArtificial Analysis Terminus 2	27.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#121	Gemma 4 12B (Reasoning)GoogleArtificial Analysis Terminus 2	27.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#121	Mercury 2InceptionArtificial Analysis Terminus 2	27.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#123	gpt-oss-120b (high)OpenAIArtificial Analysis Terminus 2	26.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#123	Mistral Small 3.1MistralArtificial Analysis Terminus 2	26.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#125	Qwen3.5 4B (Reasoning)QwenArtificial Analysis Terminus 2	25.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#126	Ling 2.6 FlashUnknownArtificial Analysis Terminus 2	24.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#126	Nemotron 3.5 LightningNVIDIAArtificial Analysis Terminus 2	24.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#128	Command A+CohereArtificial Analysis Terminus 2	22.8%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#129	EXAONE 4.5 33BUpstageArtificial Analysis Terminus 2	21.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#129	Qwen3.5 4B (Non-reasoning)QwenArtificial Analysis Terminus 2	21.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#129	Qwen3.5 9B (Non-reasoning)QwenArtificial Analysis Terminus 2	21.3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#132	Mistral Small 4 (Reasoning)MistralArtificial Analysis Terminus 2	21%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#133	Nemotron Cascade 2 30B A3BNVIDIAArtificial Analysis Terminus 2	20.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#133	Trinity Large ThinkingUnknownArtificial Analysis Terminus 2	20.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#135	Nova 2.0 Pro Preview (low)AmazonArtificial Analysis Terminus 2	19.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#136	DeepSeek R1 (Jan '25)DeepSeekArtificial Analysis Terminus 2	19.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#137	HyperNova 60B 2605AmazonArtificial Analysis Terminus 2	18.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#138	Nova 2.0 Pro Preview (Non-reasoning)AmazonArtificial Analysis Terminus 2	17.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#139	DeepSeek V3 (Dec '24)DeepSeekArtificial Analysis Terminus 2	16.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#140	Nova 2.0 Lite (high)AmazonArtificial Analysis Terminus 2	16.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#141	K2 Think V2UnknownArtificial Analysis Terminus 2	15%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#142	DeepSeek V3 0324DeepSeekArtificial Analysis Terminus 2	13.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#142	gpt-oss-120b (low)OpenAIArtificial Analysis Terminus 2	13.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#142	gpt-oss-20b (high)OpenAIArtificial Analysis Terminus 2	13.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#142	Mistral Medium 3.1MistralArtificial Analysis Terminus 2	13.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#146	DiffusionGemma 26B A4BGoogleArtificial Analysis Terminus 2	12.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#146	Magistral Medium 1.2MistralArtificial Analysis Terminus 2	12.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#148	Mistral Large 3MistralArtificial Analysis Terminus 2	12%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#148	Qwen3 235B A22B 2507 (Reasoning)QwenArtificial Analysis Terminus 2	12%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#148	Solar Pro 3UpstageArtificial Analysis Terminus 2	12%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#151	Celeris-1UnknownArtificial Analysis Terminus 2	11.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#152	Claude 3.5 HaikuAnthropicArtificial Analysis Terminus 2	10.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#152	GPT-4.1 miniOpenAIArtificial Analysis Terminus 2	10.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#154	Ministral 3 14BMistralArtificial Analysis Terminus 2	9.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#155	Llama 4 MaverickMetaArtificial Analysis Terminus 2	7.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#156	Nemotron 3 Nano Omni 30B A3B ReasoningNVIDIAArtificial Analysis Terminus 2	6.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#156	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)NVIDIAArtificial Analysis Terminus 2	6.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#156	Qwen3 Next 80B A3B (Reasoning)QwenArtificial Analysis Terminus 2	6.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#159	G9v3-3BUnknownArtificial Analysis Terminus 2	6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#160	GPT-4o miniOpenAIArtificial Analysis Terminus 2	5.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#160	Mistral Small 3.2MistralArtificial Analysis Terminus 2	5.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#162	Qwen3 32B (Reasoning)QwenArtificial Analysis Terminus 2	5.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#163	Llama 3.3 Instruct 70BMetaArtificial Analysis Terminus 2	4.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#163	Qwen3 14B (Reasoning)QwenArtificial Analysis Terminus 2	4.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#165	Gemma 3 27B InstructGoogleArtificial Analysis Terminus 2	4.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#165	Magistral Small 1.2MistralArtificial Analysis Terminus 2	4.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#165	o3-mini (high)OpenAIArtificial Analysis Terminus 2	4.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#168	Ministral 3 8BMistralArtificial Analysis Terminus 2	4.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#169	GPT-4.1 nanoOpenAIArtificial Analysis Terminus 2	3.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#169	GPT-5 mini (high)OpenAIArtificial Analysis Terminus 2	3.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#169	Llama 4 ScoutMetaArtificial Analysis Terminus 2	3.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#169	NVIDIA Nemotron 3 Nano 4BNVIDIAArtificial Analysis Terminus 2	3.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#173	Granite 4.1 8BIBMArtificial Analysis Terminus 2	3.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#174	Qwen3.5 2B (Reasoning)QwenArtificial Analysis Terminus 2	3%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#175	Granite 4.1 30BIBMArtificial Analysis Terminus 2	2.6%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#176	Qwen3 8B (Reasoning)QwenArtificial Analysis Terminus 2	2.2%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#177	Gemma 4 E4B (Reasoning)GoogleArtificial Analysis Terminus 2	1.9%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#178	Llama 3.1 Instruct 8BMetaArtificial Analysis Terminus 2	1.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#178	Qwen3 30B A3B 2507 (Reasoning)QwenArtificial Analysis Terminus 2	1.5%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#180	Granite 4.1 3BIBMArtificial Analysis Terminus 2	1.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#180	Nanbeige4.1-3BUnknownArtificial Analysis Terminus 2	1.1%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#182	Gemma 3n E4B InstructGoogleArtificial Analysis Terminus 2	0.7%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#183	Gemma 3 4B InstructGoogleArtificial Analysis Terminus 2	0.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#183	Gemma 4 E2B (Reasoning)GoogleArtificial Analysis Terminus 2	0.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#183	Phi-4 Mini InstructMicrosoftArtificial Analysis Terminus 2	0.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#183	Qwen3.5 0.8B (Non-reasoning)QwenArtificial Analysis Terminus 2	0.4%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#187	Gemma 3 12B InstructGoogleArtificial Analysis Terminus 2	0%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#187	MiniCPM-V 4.6 1.3BUnknownArtificial Analysis Terminus 2	0%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#187	Ministral 3 3BMistralArtificial Analysis Terminus 2	0%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#187	Qwen3.5 0.8B (Reasoning)QwenArtificial Analysis Terminus 2	0%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—
#187	Qwen3.5 2B (Non-reasoning)QwenArtificial Analysis Terminus 2	0%	Harness: Terminus 2 (Artificial Analysis)Metric: Pass@1 · 3 repeats	—

Full Artificial Analysis Terminus 2 / Terminal-Bench v2.1 table mirrored 2026-08-11 from artificialanalysis.ai (Pass@1, 3 repeats), with Grok 4.6 (high) 88.4% inserted from the live AA table on 2026-08-14. Proprietary agent harnesses published on tbench.ai are intentionally excluded so scores stay on one evaluator.

Open Artificial Analysis Terminal-Bench v2.1

Evidence in this catalog

Where Terminal-Bench v2.1 fits

Closest verified examples we currently carry. A missing score is not a zero.

Model examples

GPT-5.6 Sol89.5% (xhigh · AA Terminus 2)
Artificial Analysis Terminus 2 pass@1 on Terminal-Bench v2.1.
Claude Opus 589.1% (max · AA Terminus 2)
Artificial Analysis Terminus 2 pass@1 on Terminal-Bench v2.1.
Claude Fable 584.6% (AA Terminus 2)
Artificial Analysis Terminus 2 pass@1 on Terminal-Bench v2.1 — not a separate provider-only figure.

Harness examples

Claude CodeTerminal agent
Terminal-Bench scores models inside a harness loop — compare terminal agents after you narrow the model field.
CursorIDE agent harness
Use harness comparisons to separate editor workflow from bare API scores on this board.

Open this board on Artificial Analysis Terminal-Bench v2.1