Amplifying/agent-intelligence
/coding-agentsData as of Jul 21, 2026

Coding Agents IndexPreview

Index of what coding agents are doing in public repos: PRs opened, commits attributed, merge outcomes, and review patterns.

New · Research reportState of the Coding Agent Market, July 2026
TrendsWhich agents are gaining visible PR share?21.7% of all public PRs, from 8.5% six months agoCompareWhich agent PRs actually merge?Codex 86.8%, Jules 37.2% in settled PRsAutonomyHow much merged agent work had a human reviewer?54.9% reviewed, 45.1% unreviewedModel MixWhich Claude Code model appears in public runs?Opus 70.7% of observed Claude Code runsSegmentsWhere does agent work concentrate?Languages, frameworks, industries, and public ownersCoverageWhich agents can't be measured from public GitHub at all?11 of 33 leave no signal we can detect yet

Rankings

Filter by surface, pricing, security posture, and model support. Open any row for sources and caveats.

29 / 35
i
↗
1
Claude Codeest.

Anthropic

2.3M4.2M18M84M-44.1%$2.5B+rev
10.4M*
$20-$200Retail token equivalent
CLIIDEDesktopWebCloudMobile
2
OpenAI Codexest.

OpenAI

608K5.1M+2.8%n/a31.3K5M+ weeklyusers
12.2M*
$20-$200Token rate card
CLICloudDesktopIDEWebMobile
3
Julesexact

Google

185K1.2M61.1K1.1M+1.6%n/a
n/a
$0-$249.99Task quota
CloudWebCLI
4
Cursorest.

Anysphere

182K614K2.5M8.5M-35%$4Brev
n/a
$20-$200Credit proxy
IDECLICloudWeb
5
GitHub Copilotexact

GitHub / Microsoft

64.4K1.9M98.5K3.9M+0.5%4.7M paidusers
1.1M*
$10-$100Token rate card
IDECLICloudCode reviewWebMobile
6
Devinexact

Cognition

30.7K161K14.8K259K+2.6%$492Mrev
n/a
$20-$200Credit proxy
CloudWebCLIIDECode review
7
ZCode

Z.ai (Zhipu AI)

133n/a12No signal yetn/a
n/a
n/an/a
DesktopCLICloud
8
Aiderest.

Aider AI

n/an/a534613K+1.2%n/a
n/a
$0BYOK/API retail
CLI
9
Gemini CLI / Antigravity CLIest.

Google

n/an/a7K227K-11.1%n/a
444K*
$0-$199.99Task quota
CLI
10
Qwen Codeest.

Alibaba / Qwen

n/an/an/a99K+0.4%n/a
101K*
$0BYOK/API retail
CLIIDE
Common comparisonsClaude Code vs CodexCursor vs CopilotClaude Code vs CursorCodex vs CursorDevin vs Claude CodeBuild your own

Business Evidence

Public market signals behind the agent table.

Help us make this better

Email team@amplifying.ai with ideas, data, a missing agent, a correction, or anything we got wrong or left out. Or read the methodology for how every number is measured.

Reading this with an agent? There is a machine-readable Markdown version of the full index.