AI TOOL COMPARISONSCatalog snapshot

Compare AI.
Build your stack.

Compare models, coding agents and the tools they work with. See what each product supports, where it runs, and what it costs.

Source-backed specs and pricing

Sources and methodology

Reference Build

The #1 in every ranking, put together.

Build your own

Updated

The meta
coding stack.

Each slot is the #1 in its shortlist ranking, by the measure shown on its tile.

Leaders are ranked independently, so this is a starting point rather than a tested combination. Use Build your own to check compatibility and cost.

Build evidence ↗

What changed

Full change log ↗

The shortlist

AI models

Rank / ProductTerminal-Bench 4.1
  • 1
    Claude Fable 5.2
    65.8%
  • 1
    Claude Opus 5.5
    65.8%
  • 3
    Gemini 4 Carbon
    64.5%
  • 4
    Claude Sonnet 5.5
    63.9%
  • 5
    GPT-6.2 Astra
    62.1%
  • 6
    GPT-6.2 Sol
    61.5%
  • 7
    Gemini 4 Hydrogen
    56.4%
  • 8
    Gemini 4 Argon
    55.5%
  • 9
    GPT-6.1 Terra
    53.3%
  • 10
    Claude Haiku 5.5
    40.0%

Models ranked by independent benchmark score. Coding uses each model's best Terminal-Bench result in any agent; Overall uses the Artificial Analysis Intelligence Index at each model's highest-effort configuration. Models documented as lacking function calling are excluded; models whose function-calling support is not yet recorded are ranked and marked unverified. Models without a current score are listed as not benchmarked. · Source and methodology · Benchmark data: Terminal-Bench 4.1 leaderboard, measured 2026-11-19

Compare ai models ↗

Editors

Rank / ProductReported regular use
  • 1
    Visual Studio Code
    70.9%
  • 2
    Visual Studio
    27.6%
  • 3
    Vim
    27.2%
  • 4
    IntelliJ IDEA
    24.3%
  • 5
    Cursor
    14.9%
  • 6
    Neovim
    14.7%
  • 7
    PyCharm
    13.5%
  • 8
    Zed
    11.4%
  • 9
    WebStorm
    6.7%

2026 Stack Overflow Developer Survey · respondents could select multiple editors · Source and methodology · ODbL 1.0

Compare editors ↗

Coding agents

Rank / ProductReported use
  • 1
    Claude Code
    65.5%
  • 2
    Codex
    29.5%
  • 3
    Cursor
    25.5%
  • 4
    Cline
    4.5%
  • 5
    Aider
    2.1%
  • —
    Antigravity CLI
    No data
  • —
    GitHub Copilot CLI
    No data
  • —
    Grok Build
    No data
  • —
    mini-swe-agent
    No data
  • —
    Muse Code
    No data

2026 Stack Overflow Developer Survey · 12,255 respondents to the coding-agent question · Source and methodology · Benchmark data: Terminal-Bench 4.1 leaderboard, measured 2026-11-19 · ODbL 1.0

Compare coding agents ↗

Inference providers

Rank / ProductVerified model routes

Terminals

Rank / ProductHomebrew installs

Multiplexers

Rank / ProductGitHub stars

Find my AI stack

Find compatible coding tools using documented integrations. Enter an optional budget for individual software and estimated API usage, before taxes and extra charges.