Tool & model guidance

Decision Guide

Describe what you're trying to do — we'll match your task dynamically and recommend the best-fit tool and model.

How this works

  1. Describe your task in your own words — any task, not just a fixed list.
  2. We match your task against the live model registry (synced daily from Cursor and OpenAI) using Claude Haiku on AWS Bedrock.
  3. Start with the economical default — upgrade only when the result suggests you need more capability.

Which tool should I open?

ChatGPT

Text, analysis, and general knowledge work in the browser.

Use when

  • Summaries, meeting notes, and documentation drafts
  • Explaining concepts or reviewing written content
  • Light analysis, brainstorming, and stakeholder comms

Try another tool when

  • Large in-repo code edits — use Cursor instead
  • Deep infra / CI / multi-service debugging — use Codex instead
Cursor

In-editor coding — changes inside your repository.

Use when

  • Debugging or refactoring code you already have open
  • Implementing features across files in the IDE
  • Quick fixes, renames, and component-level work

Try another tool when

  • Long-form writing or policy docs — ChatGPT is usually faster
  • Terraform pipelines or CI infra — Codex is a better fit
Codex

Backend, infra, and heavier engineering tasks outside the IDE flow.

Use when

  • CI failures, flaky tests, and build pipeline issues
  • Terraform, IAM, networking, and platform modules
  • Cross-service API work and complex backend changes

Try another tool when

  • Simple prose or email drafts — ChatGPT is cheaper
  • Small single-file UI tweaks — Cursor in the IDE is faster

Want to compare models side by side or check status?

Browse AI Models →