AI Code Review, Start Now
Extracts only the essentials from your changes for precise reviews.
Claude, GPT, Gemini — pick the model that fits.
Claude Code Plugin
RecommendedOne-click install, no API key required
Code Review with an Edge!
Analyze code changes and receive quality improvement suggestions with a simple command
What Makes Selvage Special
Cutting-edge AI technology meets developer-friendly workflows to deliver a completely new code review experience
MCP Server Mode
Register as MCP server in Cursor, Claude Code, etc. and request code reviews through natural language
Agent-Delegated Review
Host agents perform code reviews with their own LLM using structured context — no API key required
Multiple AI Model Support
Leverage latest LLM models including OpenAI GPT-5, Anthropic Claude, Google Gemini
Git Workflow Integration
Supports analysis of staged, unstaged, and specific commit/branch changes
Smart Context
Provides optimized context based on Tree-sitter AST analysis
Large Context Review
Stable large-scale code review support even when exceeding model context limits
Comprehensive Code Review
Detects bugs and logic errors, suggests code quality and readability improvements
Open Source
Free to use and modify under Apache-2.0 license
Smart Context Feature
Selvage uses AST to precisely extract only the code blocks related to changed lines, ensuring both cost efficiency and review quality.
How Smart Context Works
Extracts only the minimal function/class blocks containing changed lines and related dependencies (e.g., import statements)
Significantly reduces token usage by sending only necessary context instead of entire files
Maintains high review accuracy through AST-based precise code structure understanding
Smart Context Auto Application
Selvage analyzes file size and change scope to automatically select the most efficient review method:
Auto Optimization: The optimal analysis method for each situation is automatically applied without additional configuration.
Smart Context Supported Languages
General Context Extraction Supported Languages
Supports general context extraction for major programming languages
Provides excellent code review quality in major programming languages through general context extraction methods.
Smart Context supported languages are continuously expanding.
Latest AI Models Support
Easily use various cutting-edge AI models with a single API key through OpenRouter integration
Manage all models below with one OpenRouter API key!
Access all AI models with a single API key through OpenRouter. Individual Provider API keys are also supported.
OpenAI Models(OpenRouter or OpenAI API Key)
gpt-5.2-codex
RecommendedHigh reasoning codex model (400K context)
Anthropic Models(OpenRouter or Anthropic API Key)
claude-opus-4.5
Most capable model with extended thinking (200K context)
claude-sonnet-4.5
RecommendedHigh quality reasoning model (200K context)
claude-sonnet-4-thinking
RecommendedExtended thinking process support (200K context)
claude-sonnet-4
Hybrid reasoning model optimized for coding (200K context)
Google Models(OpenRouter or Google API Key)
gemini-3-pro
Large context and advanced reasoning (1M+ tokens)
gemini-3-flash
RecommendedOptimized for response speed and cost efficiency (1M+ tokens)
OpenRouter Provided Models(OpenRouter API Key Required)
qwen3-coder
QwenRecommended480B parameter MoE coding-specialized model (1M+ tokens)
kimi-k2
Moonshot AI1T parameter MoE large-scale reasoning model (128K tokens)
minimax-m2.1
MiniMaxAdvanced reasoning with 16K thinking tokens
glm-4.7
Zhipu AIThinking-enabled reasoning model (16K thinking tokens)