ModelCompare
Compare

AI coding agent usage

What was the most used AI coding agent on ?

Claude Code from Anthropic, with a usage score of 52 out of 100 across 3 of 3 sources. Gemini CLI was next at 39.

Usage score over time, week by week

Usage ranking

AI coding agents by usage on January 1, 2026

This ranks how much each agent is used and followed, not how good it is: a popular agent can be a worse fit for your work than a small one. Click an agent to see where its numbers come from. Each source's column shows the agent's scaled score (the leader is 100) over the raw number. 3 sources had data on this day. The score averages an agent's scaled values over the sources that can measure it (at least three).

  1. 1
    Claude Code
    Anthropic · closed source
    Usage score 52
    • npm 100 · 1.7M
    • GitHub 52 · 44k
    • VS Code 4 · 2.1M
    • npm: 1001,696,092 downloads of @anthropic-ai/claude-code in the 7 days to Jan 1, 2026. That's the most of any agent in this source, so it scores 100. Source
    • GitHub: 5244,087 stars on github.com/anthropics/claude-code on Dec 1, 2025. The leader here is Gemini CLI with 85k, so Claude Code scores 44k ÷ 85k × 100 = 52. Source
    • VS Code: 42,058,732 installs of the anthropic.claude-code extension on Dec 3, 2025. The leader here is GitHub Copilot with 52M, so Claude Code scores 2.1M ÷ 52M × 100 = 4. Source

    Usage score: (100 + 52 + 4) ÷ 3 = 52. Claude Code website

  2. 2
    Gemini CLI
    Google · open source
    Usage score 39
    • npm 18 · 298k
    • GitHub 100 · 85k
    • VS Code 1 · 392k
    • npm: 18298,003 downloads of @google/gemini-cli in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Gemini CLI scores 298k ÷ 1.7M × 100 = 18. Source
    • GitHub: 10085,477 stars on github.com/google-gemini/gemini-cli on Dec 2, 2025. That's the most of any agent in this source, so it scores 100. Source
    • VS Code: 1392,333 installs of the Google.gemini-cli-vscode-ide-companion extension on Dec 12, 2025. The leader here is GitHub Copilot with 52M, so Gemini CLI scores 392k ÷ 52M × 100 = 1. Source

    Usage score: (18 + 100 + 1) ÷ 3 = 39. Gemini CLI website

  3. 3
    GitHub Copilot
    GitHub · closed source
    Usage score 34
    • npm 2 · 37k
    • VS Code 100 · 52M
    • npm: 237,111 downloads of @github/copilot in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so GitHub Copilot scores 37k ÷ 1.7M × 100 = 2. Source
    • VS Code: 10052,002,756 installs of the GitHub.copilot-chat extension on Dec 8, 2025. That's the most of any agent in this source, so it scores 100. Source

    Not counted: GitHub, because it has no public GitHub repo.

    Usage score: (2 + 100) ÷ 3 = 34, divided by 3 rather than 2 because every agent's average counts at least 3 sources. GitHub Copilot website

  4. 4
    Codex CLI
    OpenAI · open source
    Usage score 29
    • npm 18 · 305k
    • GitHub 64 · 55k
    • VS Code 5 · 2.6M
    • npm: 18304,664 downloads of @openai/codex in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Codex CLI scores 305k ÷ 1.7M × 100 = 18. Source
    • GitHub: 6455,092 stars on github.com/openai/codex on Jan 2, 2026. The leader here is Gemini CLI with 85k, so Codex CLI scores 55k ÷ 85k × 100 = 64. Source
    • VS Code: 52,572,471 installs of the openai.chatgpt extension on Dec 13, 2025. The leader here is GitHub Copilot with 52M, so Codex CLI scores 2.6M ÷ 52M × 100 = 5. Source

    Usage score: (18 + 64 + 5) ÷ 3 = 29. Codex CLI website

  5. 5
    Cline
    Cline · open source
    Usage score 24
    • npm 1 · 9.5k
    • GitHub 66 · 57k
    • VS Code 5 · 2.8M
    • npm: 19,487 downloads of cline in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Cline scores 9.5k ÷ 1.7M × 100 = 1. Source
    • GitHub: 6656,576 stars on github.com/cline/cline on Jan 1, 2026. The leader here is Gemini CLI with 85k, so Cline scores 57k ÷ 85k × 100 = 66. Source
    • VS Code: 52,812,341 installs of the saoudrizwan.claude-dev extension on Dec 29, 2025. The leader here is GitHub Copilot with 52M, so Cline scores 2.8M ÷ 52M × 100 = 5. Source

    Usage score: (1 + 66 + 5) ÷ 3 = 24. Cline website

  6. 6
    OpenCode
    Anomaly · open source
    Usage score 20
    • npm 6 · 93k
    • GitHub 53 · 46k
    • VS Code 0 · 18k
    • npm: 693,344 downloads of opencode-ai in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so OpenCode scores 93k ÷ 1.7M × 100 = 6. Source
    • GitHub: 5345,642 stars on github.com/anomalyco/opencode on Jan 2, 2026. The leader here is Gemini CLI with 85k, so OpenCode scores 46k ÷ 85k × 100 = 53. Source
    • VS Code: 018,257 installs of the sst-dev.opencode extension on Nov 11, 2025. The leader here is GitHub Copilot with 52M, so OpenCode scores 18k ÷ 52M × 100 = 0. Source

    Usage score: (6 + 53 + 0) ÷ 3 = 20. OpenCode website

  7. 7
    Roo Code
    Roo Code · open source
    Usage score 9
    • GitHub 25 · 22k
    • VS Code 2 · 1M
    • GitHub: 2521,540 stars on github.com/RooCodeInc/Roo-Code on Jan 2, 2026. The leader here is Gemini CLI with 85k, so Roo Code scores 22k ÷ 85k × 100 = 25. Source
    • VS Code: 21,048,984 installs of the RooVeterinaryInc.roo-cline extension on Dec 4, 2025. The leader here is GitHub Copilot with 52M, so Roo Code scores 1M ÷ 52M × 100 = 2. Source

    Not counted: npm, because it has no command-line package on npm.

    Usage score: (25 + 2) ÷ 3 = 9, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Roo Code website

  8. 8
    Qwen Code
    Alibaba · open source
    Usage score 7
    • npm 1 · 24k
    • GitHub 19 · 16k
    • npm: 123,523 downloads of @qwen-code/qwen-code in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Qwen Code scores 24k ÷ 1.7M × 100 = 1. Source
    • GitHub: 1916,018 stars on github.com/QwenLM/qwen-code on Dec 3, 2025. The leader here is Gemini CLI with 85k, so Qwen Code scores 16k ÷ 85k × 100 = 19. Source

    Not counted: VS Code, because it has no VS Code extension.

    Usage score: (1 + 19) ÷ 3 = 7, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Qwen Code website

  9. 9
    Crush
    Charm · open source
    Usage score 7
    • npm 0 · 2k
    • GitHub 19 · 17k
    • npm: 01,952 downloads of @charmland/crush in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Crush scores 2k ÷ 1.7M × 100 = 0. Source
    • GitHub: 1916,570 stars on github.com/charmbracelet/crush on Jan 1, 2026. The leader here is Gemini CLI with 85k, so Crush scores 17k ÷ 85k × 100 = 19. Source

    Not counted: VS Code, because it has no VS Code extension.

    Usage score: (0 + 19) ÷ 3 = 7, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Crush website

  10. 10
    Kilo Code
    Kilo · open source
    Usage score 6
    • npm 0 · 4.2k
    • GitHub 16 · 14k
    • VS Code 1 · 561k
    • npm: 04,235 downloads of @kilocode/cli in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Kilo Code scores 4.2k ÷ 1.7M × 100 = 0. Source
    • GitHub: 1613,662 stars on github.com/Kilo-Org/kilocode on Jan 1, 2026. The leader here is Gemini CLI with 85k, so Kilo Code scores 14k ÷ 85k × 100 = 16. Source
    • VS Code: 1560,586 installs of the kilocode.Kilo-Code extension on Dec 16, 2025. The leader here is GitHub Copilot with 52M, so Kilo Code scores 561k ÷ 52M × 100 = 1. Source

    Usage score: (0 + 16 + 1) ÷ 3 = 6. Kilo Code website

  11. 11
    Auggie
    Augment Code · closed source
    Usage score 1
    • npm 1 · 9.2k
    • VS Code 1 · 649k
    • npm: 19,231 downloads of @augmentcode/auggie in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Auggie scores 9.2k ÷ 1.7M × 100 = 1. Source
    • VS Code: 1648,968 installs of the augment.vscode-augment extension on Dec 6, 2025. The leader here is GitHub Copilot with 52M, so Auggie scores 649k ÷ 52M × 100 = 1. Source

    Not counted: GitHub, because it has no public GitHub repo.

    Usage score: (1 + 1) ÷ 3 = 1, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Auggie website

  12. 12
    Amp
    Sourcegraph · closed source
    Usage score 0
    • npm 1 · 20k
    • VS Code 0 · 69k
    • npm: 120,397 downloads of @sourcegraph/amp in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Amp scores 20k ÷ 1.7M × 100 = 1. Source
    • VS Code: 068,667 installs of the sourcegraph.amp extension on Dec 3, 2025. The leader here is GitHub Copilot with 52M, so Amp scores 69k ÷ 52M × 100 = 0. Source

    Not counted: GitHub, because it has no public GitHub repo.

    Usage score: (1 + 0) ÷ 3 = 0, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Amp website

  13. 13
    pi
    Mario Zechner · open source
    Usage score 0
    • npm 0 · 2.6k
    • GitHub 1 · 888
    • npm: 02,580 downloads of @mariozechner/pi-coding-agent in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so pi scores 2.6k ÷ 1.7M × 100 = 0. Source
    • GitHub: 1888 stars on github.com/badlogic/pi-mono on Jan 1, 2026. The leader here is Gemini CLI with 85k, so pi scores 888 ÷ 85k × 100 = 1. Source

    Not counted: VS Code, because it has no VS Code extension.

    Usage score: (0 + 1) ÷ 3 = 0, divided by 3 rather than 2 because every agent's average counts at least 3 sources. pi website

  14. 14
    Letta Code
    Letta · open source
    Usage score 0
    • npm 0 · 1.7k
    • GitHub 0 · 159
    • npm: 01,663 downloads of @letta-ai/letta-code in the 7 days to Jan 1, 2026. The leader here is Claude Code with 1.7M, so Letta Code scores 1.7k ÷ 1.7M × 100 = 0. Source
    • GitHub: 0159 stars on github.com/letta-ai/letta-code on Dec 16, 2025. The leader here is Gemini CLI with 85k, so Letta Code scores 159 ÷ 85k × 100 = 0. Source

    Not counted: VS Code, because it has no VS Code extension.

    Usage score: (0 + 0) ÷ 3 = 0, divided by 3 rather than 2 because every agent's average counts at least 3 sources. Letta Code website

Not ranked that day, only one source has them: OpenHands (GitHub, 66k), Aider (GitHub, 39k), Droid (npm, 3).

Method

How the usage score is made

No single public number measures how much each AI coding agent is used, so we blend the ones that exist. In each source the leader scores 100 and the rest are scaled to it. An agent's score is the average over the sources that can measure it. Sources that can't see an agent at all (Cursor has no npm package) are left out of its average, but the average always counts at least three sources, so an agent seen by only one or two can't outrank one that leads everywhere. An agent needs two sources to be ranked.

  • npmCLI downloads, 7 days · data from Jan 1, 2026

    Downloads of each agent's command-line package in the 7 days to the date, with one-off bursts (mirrors, bots) capped. Downloads count installs and CI runs, not people, and agents with their own installer are undercounted. Source

  • GitHubStars · data from Jan 2, 2026

    Stars on the agent's GitHub repo, from monthly snapshots. Stars measure interest, and build up over time. Source

  • VS CodeExtension installs, all time · data from Dec 29, 2025

    All-time installs of the agent's VS Code extension, from monthly snapshots. Cursor and Windsurf install from Open VSX instead, which isn't counted. Source

Browse

The most used AI coding agent, month by month

2026
2025

Pick the model behind your agent

Most of these agents let you choose the model. Send one prompt to up to six models and see which answer you'd actually use, with what each one cost.

Compare models

Data through October 21, 2026 for 20 agents. npm downloads from the npm registry; OpenRouter app rankings from OpenRouter (CC BY 4.0); GitHub stars and VS Code installs from monthly snapshots in the Internet Archive and the live APIs; survey figures as published by Stack Overflow and JetBrains.