Open-source AI code reviewer
Scopy is an AI-powered code reviewer that understands your repository, catches bugs and improves code quality.
How code review works with Scopy AI
Connect GitHub, open a pull request and get focused, verified feedback without changing how your team already works.
Connect your repository
Install the Scopy GitHub App and choose which repositories should get automated code reviews. Setup takes under ten minutes.
Open a pull request
Scopy reviews new pull requests automatically, building repository-wide context around the diff before deciding what to flag.
Get verified feedback
An independent pass challenges each candidate issue before focused comments land on the exact lines that matter.
Accuracy comes from the process
More context for finding real bugs. More scrutiny before reporting them.
Scopy traces changes through surrounding code to understand intent, surface area and downstream risks before it writes a word.
Bring your own key on hosted platform, or self-host with any compatible API. No model lock-in.
Turn your engineering standards into plain-language checks that run consistently on every change.
Scopy detects the libraries your code uses and checks findings against their official documentation, citing sources instead of guessing API behavior.
Built in the open
Inspect every line, fork the project or run Scopy on your own infrastructure.
Self-host
- MIT licensed, full source code
- Connect any LLM provider
- Your data stays in your environment
- Community support
Use Scopy AI in cloud
- Connect repositories in seconds
- Review credits included in plan
- Unlimited repositories
- Team workspace management
Simple, credit-based pricing
Reviews use credits based on how much a pull request changes – no per-seat pricing.
Self-hosted version with any LLM provider
Start free, then add a paid plan for managed cloud reviews – or self-host without limits.
- Bring-your-own-key support
- Self-host with your own model keys
- Full source code access
25 credits / month included
Hosted Scopy for teams that ship fast.
- Unlimited repositories
- Team workspace management
- Email support
150 credits / month included
More monthly credits for teams that value quality
- Everything in Premium
- Higher monthly review credits
- Priority support
Tailored credits for your organization
For teams with custom requirements, higher volume, or dedicated support needs.
- Custom credits & pricing
- Dedicated onboarding
- Priority support & SLAs
Frequently asked questions
The basics on access, reviews, billing and deployment.
Scopy builds context from the change, affected symbols and the wider repository, then investigates potential bugs and independently verifies candidate findings before returning focused feedback.
Reviews run automatically for enabled repositories when relevant pull request activity arrives from GitHub, such as a new pull request or a draft PR being marked ready for review. You can also request a fresh review by mentioning Scopy in a PR comment.
Scopy uses a GitHub App installation to receive webhook events, read repository metadata and code for selected repositories, and publish pull request feedback. GitHub controls which repositories are visible to the app, and Scopy can only review repositories granted to that installation.
Billing is managed per workspace. Premium and Ultra are monthly plans with included review credits. Reviews consume credits based on reviewable changed lines after repository path filters: most normal pull requests use 1 credit, while larger pull requests use more.
Yes. Any workspace can switch to bring-your-own-key and add an OpenRouter or Vercel AI Gateway key. Those reviews run on your key and are billed to your provider account instead of workspace credits, and you can override the choice per repository. Keys are encrypted at rest, and self-hosted instances can use their own keys too.
Yes. Scopy AI is MIT licensed and the source code is available on GitHub. Self-hosting lets you run Scopy AI on your own infrastructure and connect your preferred model provider.
Yes. Scopy detects libraries from your dependency manifests and keeps an indexed copy of their official documentation. During review it can search those docs and answer focused questions with citations, so verdicts about API behavior rest on documented facts rather than model memory. Teams can also add custom documentation sources for internal SDKs or custom libraries.
Yes. You can configure repositories and review criteria so Scopy focuses on the rules and risks that matter for your team, including custom linting guidance and review settings.