Browse 435 LLMs

A long-horizon repository benchmark built to test realistic software engineering work. Its scores need a task-quality and setup check before they support a coding-agent decision.

1. 1
   
   Claude Fable 5.181.2
2. 2
   
   Claude Mythos 580.3
3. 3
   
   Claude Fable 580.0

73 models

A multilingual software-engineering benchmark for real-world code issue resolution across multiple programming languages.

1. 1
   
   Claude Opus 589.5
2. 2
   
   Claude Fable 5.189.1
3. 3
   
   Claude Opus 4.884.4

44 models

An independently evaluated Tau3 banking benchmark from Artificial Analysis.

1. 1
   
   Grok 4.650.7
2. 2
   
   Muse Spark 1.350.5
3. 3
   
   GLM-5.350.3

15 models

How important are skills for agents?

1. 1
   
   DeepSeek V4.1 Flash69.8
2. 2
   
   Grok 4.566.0
3. 3
   
   Gemini 3.7 Flash65.9

35 models

A broad research-agent benchmark for open-ended information gathering, synthesis, and answer construction across wide search spaces.

1. 1
   Me
   Mercury 292.3
2. 2
   Hy
   Hy4 preview83.9
3. 3
   
   Qwen3.8 Max81.9

16 models

A benchmark for web-browsing agents that must search, inspect sources, gather evidence, and return the correct answer to research-oriented questions.

1. 1
   At
   Atria Dawn Preview92.5
2. 2
   
   GPT-5.6 Sol92.2
3. 3
   
   GPT-6 Astra91.5

45 models

MRCR v2 slice focused on very long contexts at 128K-256K lengths.

1. 1
   
   GPT-5.587.5
2. 2
   
   Claude Opus 4.7 (Adaptive)59.2

2 models

Human-validated software engineering issues from real-world Python repositories.

1. 1
   
   Claude Opus 4.7 (max)83.5
2. 2
   
   GPT-5.5 (xhigh)80.6
3. 3
   
   Gemini 3.5 Flash (high)79.3

33 models

Factoid questions spanning politics, science, technology, art, sports, geography, and music.

1. 1
   
   GPT-6 Astra (max)75.6
2. 2
   
   Gemini 3.1 Pro Preview (high)73.5
3. 3
   
   Claude Fable 5.1 (max)70.8

80 models

Abstract reasoning and pattern generalization on grid-based tasks.

1. 1
   
   GPT-6 Astra (max)95.0
2. 2
   
   GPT-6 Astra (xhigh)93.3
3. 3
   
   GPT-5.6 Sol (max)92.5

203 models

Exceptionally difficult research-level mathematics problems.

1. 1
   
   GPT-6 Astra (high)97.6
2. 2
   
   GPT-6 Astra (xhigh)97.6
3. 3
   
   GPT-6 Astra (max)97.6

62 models

Real-world multi-step tool-use evaluation through the Model Context Protocol.

1. 1
   
   Muse Spark 1.188.1
2. 2
   Fa
   Fable 5.187.2
3. 3
   
   claude-opus-5 (xhigh)85.8

35 models

Real-world cybersecurity evaluation of AI agents reproducing vulnerabilities with working proof-of-concept tests.

1. 1
   
   Multi-model (Creation Model, DeepSeek-V4-Pro, Qwen 3.8 Max)98.5
2. 2
   
   GLM-5.397.2
3. 3
   
   GLM-5.396.8

70 models

The official leaderboard for Terminal-Bench 3.0.

1. 1
   Op
   Opus 542.7
2. 2
   
   GPT-5.6 Sol34.6
3. 3
   Fa
   Fable 534.0

12 models

Expert-level questions across mathematics, science, and humanities, as published by the CAIS AI Dashboard.

1. 1
   Fa
   Fable 5.154.6
2. 2
   
   GPT-6 Astra53.6
3. 3
   Fa
   Fable 552.7

60 models

Average of the text capability benchmarks published by the CAIS AI Dashboard.

1. 1
   
   GPT-6 Astra63.9
2. 2
   Fa
   Fable 5.155.6
3. 3
   Fa
   Fable 554.4

60 models

Average of the vision capability benchmarks published by the CAIS AI Dashboard.

1. 1
   
   GPT-6 Astra83.0
2. 2
   
   Gemini 3.8 Flash69.9
3. 3
   Fa
   Fable 5.168.9

48 models

Average of the risk and safety benchmarks published by the CAIS AI Dashboard; lower is better.

1. 1
   Fa
   Fable 5.130.3
2. 2
   
   Muse Spark 1.132.5
3. 3
   
   GPT-6 Astra35.5

9 models

Remote Labor Index automation rates published by the CAIS AI Dashboard.

1. 1
   Fa
   Fable 515.8
2. 2
   Op
   Opus 4.88.3
3. 3
   
   GPT-5.56.3

13 models

Composite Artificial Analysis index across mathematics, science, coding, and reasoning evaluations.

1. 1
   
   Claude Fable 5.1 (max with fallback)53.4
2. 2
   
   Claude Fable 5.1 (xhigh with fallback)53.2
3. 3
   
   GPT-6 Astra (max)52.8

633 models

Artificial Analysis composite index for programming and software-engineering capability.

1. 1
   
   Claude Fable 5.1 (max with fallback)81.6
2. 2
   
   Claude Fable 5.1 (xhigh with fallback)80.7
3. 3
   
   Claude Fable 5.1 (high with fallback)79.1

256 models

Artificial Analysis composite index for tool use, planning, autonomy, and complex agentic workflows.

1. 1
   
   Claude Fable 5.1 (max with fallback)58.0
2. 2
   
   Claude Fable 5.1 (xhigh with fallback)57.1
3. 3
   
   Claude Opus 5 (max)56.2

151 models

Artificial Analysis factual-knowledge and hallucination-resistance evaluation.

1. 1
   
   GPT-6 Astra (high)43.7
2. 2
   
   Claude Fable 5.1 (max with fallback)43.5
3. 3
   
   GPT-6 Astra (xhigh)43.4

99 models

Graduate-level, expert-written science questions evaluated by Artificial Analysis.

1. 1
   
   GPT-6 Astra (xhigh)96.3
2. 2
   
   GPT-6 Astra (max)96.1
3. 3
   
   Gemini 3.8 Flash (high)95.3

612 models

Broad expert-level reasoning and knowledge evaluation published by Artificial Analysis.

1. 1
   
   Claude Fable 5.1 (max with fallback)59.1
2. 2
   
   Claude Fable 5.1 (xhigh with fallback)58.7
3. 3
   
   Claude Fable 5.1 (high with fallback)55.9

607 models

Instruction-following benchmark results published by Artificial Analysis.

1. 1
   
   Grok 4.3 (medium)83.3
2. 2
   
   Grok 4.20 030982.9
3. 3
   
   MiniMax-M382.9

450 models

Scientific programming benchmark results published by Artificial Analysis.

1. 1
   
   Claude Fable 5.1 (max with fallback)63.1
2. 2
   
   Claude Fable 5 (with fallback)61.0
3. 3
   
   Claude Fable 5.1 (xhigh with fallback)60.9

167 models

Agentic terminal and software-engineering results published by Artificial Analysis.

1. 1
   
   Claude Fable 5.1 (max with fallback)91.4
2. 2
   
   Claude Fable 5.1 (xhigh with fallback)91.0
3. 3
   
   Claude Fable 5.1 (high with fallback)89.9

236 models

Critical-points programming evaluation published by Artificial Analysis.

1. 1
   
   GPT-5.6 Sol (max)32.3
2. 2
   
   GPT-6 Astra (max)31.7
3. 3
   
   GPT-6 Astra (xhigh)31.4

520 models

Long-context reasoning results published by Artificial Analysis.

1. 1
   
   Kimi K3 (max)88.7
2. 2
   
   Claude Fable 5.1 (max with fallback)85.3
3. 3
   
   Claude Fable 5.1 (medium with fallback)84.7

516 models

Tool-agent interaction benchmark results published by Artificial Analysis.

1. 1
   JT
   JT-35B-Flash99.1
2. 2
   
   GLM-5.2 (max)99.1
3. 3
   
   GLM-4.7-Flash98.8

440 models