Atlas

Benchmarks

Artificial Analysis Coding Agent Index v2.1

Artificial Analysis Coding Agent Index v2.1 scores coding-agent and model pairs, such as Claude Code or Codex with a given model and effort, as the equal-weight average of mean scores over three attempts per task on FrontierCode-AA 2.0 (132 tasks; each attempt scores 0 or between 70% and 100%), Terminal-Bench 4.1 (66 tasks) and SWE-Atlas-QnA-AA (124 tasks, grading adjusted by Artificial Analysis). FrontierCode-AA 2.0 replaces v2.0's FrontierCode v1.1 and Terminal-Bench 4.1 replaces Terminal-Bench 4.0, so v2.1 scores are not comparable with v2.0. Scores belong to the agent harness and model together. It is separate from the model-level Artificial Analysis Coding Index.

Models
17
Top score
55.5
Leader
Gemini 4 Carbon
Updated
—
Resolution
—
Saturation
—

No signal or difficulty estimate. Too few models have been measured on this benchmark to estimate its difficulty.

Organizations
16 dated models · 4 frontier modelsClosed weightsOpen weights
Data: Artificial Analysisatlas.village.dev

Artificial Analysis Coding Agent Index v2.1 leaderboard

#ModelSettingScoreSource
1Gemini 4 Carbonhigh · Antigravity CLI
55.5
2Muse Wonder Fullmax · Muse Code
54.5
3GPT-6 Astramax · Codex
53.1
4GPT-6.1 Solmax · Codex
52.2
5GPT-6.1 Terramax · Codex
48.6
6Kimi K3.1 Promax · Kimi Code
48.6
7Gemini 4 Argonhigh · Antigravity CLI
45.5
8Muse Wondermax · Muse Code
44.5
9Grok 4.7x-high · Grok Build
41.6
10Mistral Large 4high · Mistral Vibe
40.6
11GLM-5.3max · OpenCode
40.1
12Muse Embermax · Muse Code
39.7
13DeepSeek V4.1 Flash 1017max · DeepSeek Harness
36.8
14Muse Spark 1.3max · Muse Code
36.8
15Step 5 Previewnot reported · Claude Code
35.8
16GPT-6 Lunamax · Codex
25.9
17Beam 1.0 501B A23Bnot reported · Mirror
15.5

Related benchmarks

Loading Atlas data…