Choose your plan
Local-first and model-agnostic. Bring your own key (OpenAI, Anthropic, Gemini) or run a private local model, and pay providers directly — zero markup on AI costs. The commit loop always runs on your machine.
Solo
For individual developers and agent-heavy solo workflows.
- Unlimited commit-triggered reviews
- Bring your own key or run a local model
- The Refactor Terminal (
refactor watch) - Pre-push gate (
warn/strict) - Behavior-preserving Refactor Requests
- Runs fully local — no code leaves your machine
- Community support
Team
For teams that want shared policy and managed inference.
- Everything in Solo
- Shared team pre-push policy (
warn/strict) - Organizations, teams & project collaboration
- Managed inference option (or keep BYOK / local)
- Priority review queue
- Engineering metrics & delivery insights
- Community & email support
Enterprise
For organizations needing SSO, compliance and dedicated support.
- Everything in Team
- SSO / SAML
- RBAC + audit logs and analytics
- Self-hosted runtime & managed toolchain
- SOC 2
- Dedicated onboarding & support SLA
- Private support channel
You pay RefactorAI for the platform, and your LLM provider for API usage — we never mark up your tokens. Choose a local model and your LLM cost is $0.
You own the model — and the bill
RefactorAI is model-agnostic. The platform fee is fixed per dev; token cost is whatever your provider charges (or zero, with a local model).
Bring your own key
Use your OpenAI, Anthropic, or Gemini key and pay the provider directly at list price. Zero markup from us.
Or run it local — $0 tokens
Run a license-clean local model in an isolated container. No provider bill, and nothing ever leaves the machine.
Commit-scoped = cheaper
Reviews target changed + associated files, not the whole repo — so tokens scale with your diff, not your codebase.
Estimate your monthly cost
Platform license plus your own token spend. A local model drops LLM cost to zero.
Token estimates are illustrative and paid directly to your provider.
Pricing & billing questions
Do I pay twice if I use BYOK (RefactorAI + my provider)?
You pay RefactorAI for the platform (per dev) and your LLM provider for API usage. We don't charge any markup on your tokens. Run a local model and the token cost is $0.
Is it really free for solo developers?
Yes. The Solo plan gives you unlimited commit-triggered reviews, the Refactor Terminal, and the pre-push gate — you just bring your own key or run a local model.
Do you store or train on my source code?
No. The commit loop runs locally with no network round-trip on the commit path. Your source and index never leave your machine, and we never train on your code.
Will I be charged for every developer in my org?
No — you decide who's on the RefactorAI team and are only billed for those seats. You control team membership and billing.
Which models are supported?
RefactorAI is model-agnostic: use Claude, GPT, Gemini, or a license-clean local model. You choose the model, the cost, and the rules.
How is this different from a PR review tool?
PR tools review after you push and leave comments on a PR that already exists. RefactorAI reviews on every local git commit, fixes issues in the Refactor Terminal, and gates the push — so the PR arrives clean.
Start free. Upgrade when your team grows.
Install the CLI, review your next commit locally, and only push what passes the gate.