✨ Features
- ⚡ Generate commit messages automatically from git diff
- 🤖 Supports OpenAI / Azure OpenAI / Gemini / DeepSeek APIs
- 🌍 Supports multi-language commit messages
- 😎 Optional Gitmoji support
- 🛠️ Custom system prompt
- 📏 Follows Conventional Commits specification
- 🎨 Simple and fast UI inside the Source Control panel
- 📊 Benchmarking: View token usage and generation time in the "CommiX" output channel
📦 Installation
From VS Code Marketplace
Search for “CommiX” → Click Install
Requirements
🤯 Usage
- Install CommiX extension in VS Code
- Open Settings → CommiX → Configure your API key & model
- Stage your changes (
git add)
- Click the CommiX AI button next to the commit message input
- CommiX generates a commit message and inserts it automatically
- Review & commit
💡 If the diff is too large, stage files in smaller batches.
🛠️ AI Edit & Refactor
CommiX now supports AI-powered code editing and refactoring.
- Select the code you want to modify in the editor.
- Right-click and choose "CommiX: $(edit) AI Edit / Refactor".
- Enter your instruction (e.g., "Refactor this function to be async", "Add error handling", "Optimize for performance").
- CommiX will process your request and update the code automatically.
🚀 New in v3.0: AI Edit Mode
CommiX v3.0 introduces a powerful set of AI editing tools:
- Function-level Icons (CodeLens): "AI Edit" and "AI Explain" buttons appear directly above functions/classes.
- Inline Suggestions: Click the lightbulb 💡 icon on selected code to access AI actions.
- Convert Code: Easily convert code between languages (e.g., JS → TS) via
CommiX: Convert Code.
- Diff-based Preview: All AI edits now open in a Diff View, letting you review changes safely before applying.
- Smart Patch: Apply AI suggestions with a single click from the diff view.
⚙️ Configuration
These settings are available under:
VS Code → Settings → CommiX
| Setting |
Type |
Default |
Required |
Description |
AI_PROVIDER |
string |
openai |
Yes |
Choose: openai, gemini, nvidia |
OPENAI_API_KEY |
string |
— |
Yes (if openai) |
OpenAI API Key |
OPENAI_MODEL |
string |
gpt-4o |
Yes |
Model used |
GEMINI_API_KEY |
string |
— |
Yes (if gemini) |
Gemini API Key |
GEMINI_MODEL |
string |
gemini-2.0-flash-001 |
Yes |
Gemini model |
TEMPERATURE |
number |
0.7 |
No |
Controls creativity |
LANGUAGE |
string |
en |
Yes |
Commit message language |
SYSTEM_PROMPT |
string |
— |
No |
Custom system prompt |
ENABLE_GITMOJI |
boolean |
false |
No |
Enable Gitmoji |
NVIDIA_API_KEY |
string |
— |
Yes (if nvidia) |
NVIDIA API Key |
NVIDIA_MODEL |
string |
meta/llama-3.1-70b-instruct |
Yes |
Model used |
📘 API Key Setup Guide — CommiX
CommiX works with OpenAI, Google Gemini, and NVIDIA NIM.
Follow this guide to generate API keys for each provider.
🟦 1. OpenAI API Key
Steps:
- Open the OpenAI API dashboard:
👉 https://platform.openai.com/account/api-keys
- Sign in to your OpenAI account
- Click “Create new secret key”
- Copy the key
- In VS Code:
Settings → CommiX → OpenAI API Key
Recommended Models:
gpt-4ogpt-4o-minio3-minio1
🟩 2. Google Gemini API Key
Steps:
- Visit Google AI Studio:
👉 https://aistudio.google.com/app/apikey
- Sign in with your Google account
- Click “Create API Key”
- Copy the key
- In VS Code:
Settings → CommiX → Gemini API Key
Recommended Models:
gemini-2.0-flash-001gemini-1.5-flashgemini-1.5-pro
🟧 3. NVIDIA NIM API Key
Steps:
- Go to NVIDIA Build Platform:
👉 https://build.nvidia.com
- Sign in using your NVIDIA or Google account
- Click API Keys in the left sidebar
- Press “Create Key”
- Copy the key
- In VS Code:
Settings → CommiX → NVIDIA API Key
Recommended Models:
meta/llama-3.1-70b-instructmeta/llama-3.1-8b-instructgoogle/gemma-2-9b-itdeepseek-r1deepseek-coder-v2-lite
🟧 4. Groq API Key
Steps:
- Go to Groq Cloud Console:
👉 https://console.groq.com/keys
- Sign in
- Create an API Key
- Copy the key
- In VS Code:
Settings → CommiX → Groq API Key
Recommended Models:
llama3-70b-8192mixtral-8x7b-32768
🦙 5. Local Ollama
Steps:
- Install Ollama:
👉 https://ollama.com
- Run a model (e.g.,
ollama run llama3)
- In VS Code:
Settings → CommiX → AI Provider → ollama
- (Optional) Set
OLLAMA_BASE_URL if not using default http://localhost:11434
Recommended Models:
🔒 Security Tips
- Never commit your API keys to GitHub
- Use environment variables or VS Code settings only
- Rotate keys if exposed
- Enable billing only when required
📌 Need help?
Open an issue:
👉 https://github.com/paswans05/commix/issues
