Comments
>log in to comment
Unlike typical CI bots that only see the diff, Mira indexes your whole repository to provide context‑aware AI reviews. It catches bugs, security issues, and style drift, runs on any infrastructure with your own model, and stays open‑source. I built it to keep control while getting a second set of eyes. Would love your thoughts 😊
More AI tools →74 free dev tools →
- AI‑powered code reviews with full repository context
- Self‑hosted, open‑source under Apache 2.0
- Bring your own LLM – OpenRouter, OpenAI‑compatible or local models
- Integrates with GitHub, GitLab and Forgejo pull requests
- Detects bugs, security issues and team‑specific style violations
Open-source, self-hosted AI code reviewer for pull requests that analyzes the whole repo.
- for
- Engineering teams that want AI code reviews while keeping data on-premise.
- pricing
- open source
- license
- Apache-2.0
Key features
- Cross-file context — Indexes files, symbols and dependencies to understand changes beyond the diff.
- Bug & regression detection — Find missing awaits, edge cases and regressions before merge.
- Security review — Surface unsafe patterns and authentication gaps in PRs.
- Style & convention enforcement — Learns team standards from repository history and applies them.
- Custom model support — Use any OpenAI-compatible, OpenRouter, Ollama, vLLM or Bedrock model.
- Git provider integration — Works with GitHub, GitLab and Forgejo via apps or access tokens.
- Docker deployment — Run the full engine via Docker, Railway, Fly.io or Render.
- Open source license — Apache 2.0, free forever with community support.
Use cases
- Add an automated AI reviewer to every pull request to catch bugs early.
- Enforce security best practices and authentication checks across the codebase.
- Maintain consistent coding style by learning from existing repository conventions.
Mira Code vs alternatives
 Mira Code |  Qodo |  Augment Code |  Tabby ML |  Continue.dev | |
|---|---|---|---|---|---|
| Best for | Self-hosted AI code review with full repo context | AI code review as a service | AI coding assistant with context | Self-hosted AI code completion | IDE-integrated AI assistance |
| Pricing | Open source | Subscription | Subscription | Free | Free |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Nov 2026 | — | — | — | — |
- Mira Code vs Qodo: Qodo focuses on AI code review and test generation but is a SaaS product, not self-hosted.
- Mira Code vs Augment Code: Augment Code provides an AI coding agent with deep codebase context, not a dedicated PR reviewer.
- Mira Code vs Tabby ML: Tabby ML is a self-hosted AI coding assistant, not specialized for PR reviews.
- Mira Code vs Continue.dev: Continue.dev offers IDE-integrated AI checks and assistance, whereas Mira runs as a PR reviewer service.
Mira Code FAQ
Is Mira really free?+
Yes. Mira is open source under Apache 2.0 and can be self-hosted at no license cost.
Does my code leave my infrastructure?+
Indexes and review history stay on your infrastructure; only the context needed for a review is sent to the LLM provider you configure.
Which models can I use?+
Any OpenAI-compatible endpoint, OpenRouter, local models via Ollama or vLLM, and AWS Bedrock.
What Git providers are supported?+
Mira works with GitHub, GitLab and Forgejo (including Codeberg) via a GitHub App or access token.
How do I get started?+
Deploy Mira with Docker (or Railway, Fly.io, Render), configure your model and Git credentials, then connect your repositories.
Summarized by DevHunt from miracode.ai · Nov 11, 2026. Details may change; check the official site.
