aicharts icon
aicharts icon

aicharts

 Like

aicharts plots every current Anthropic, OpenAI and Google coding model on one chart: how much of Terminal-Bench its coding agent passed on the official leaderboard, against what that run cost per task or the output tokens it used.

Explorer: each model's Terminal-Bench score against its cost per task

Cost / License

  • Free
  • Open Source (MIT)

Since

  • New

GitHub

  • GitHub stars: 1

Platforms

  • Online
0likes
0articles
Save

Features

Features

Highlights All activities

aicharts News & Activities

Recent activities

aicharts information

GitHub repository

  •  1 Star
  •  0 Forks
  •  0 Open Issues
  •   Updated  
View on GitHub
aicharts was added to AlternativeTo by ben-8360 on and this page was last updated .

No comments or reviews, maybe you want to be first?

Official Links

What is aicharts?

aicharts plots every current Anthropic, OpenAI and Google coding model on one chart: how much of Terminal-Bench its coding agent passed on the official leaderboard, against what that run cost per task or the output tokens it used. Models that no other model beats for the money sit on the cost frontier and earn a star.

Each model has a page with its Terminal-Bench run, Epoch AI's MirrorCode, GPQA Diamond and FrontierMath scores, and its API prices. Comparison pages put two models side by side, and a coding agent page compares Codex, Claude Code and Antigravity CLI through the models run inside each.

The data refreshes several times a day from open sources: the Terminal-Bench leaderboard (Apache-2.0), Epoch AI (CC BY 4.0) and LiteLLM's price list (MIT). It is free with no sign-up, the full dataset is available as JSON through a public API, and the code is open source under MIT.