Free AI Commit Message
Generate clean, conventional Git commit messages out of the box with zero setup, or connect your own free BYOK providers (Gemini, Groq, OpenRouter) and offline local Ollama.
Works instantly upon installation with zero API key configuration, or connect directly using your own API keys to leverage generous free daily quotas from Gemini (1,500 req/day), Groq (14,400 req/day), GitHub Models, and OpenRouter, or run 100% offline with local Ollama (no key required). Also supports DeepSeek, OpenAI, and Anthropic Claude.
Using a JetBrains IDE? (Rider, IntelliJ IDEA, PyCharm, WebStorm, etc.) Check out the sibling plugin: Free AI Commit Message for JetBrains IDEs.
Quick Start (Zero Setup Required)
- Stage your Git changes:
git add <files...>
- Generate your Commit Message:
- Click the Sparkle icon (
✨) in the Source Control (SCM) title bar, or press Ctrl+Shift+P (or Cmd+Shift+P on macOS) and run Free AI Commit: Generate AI Commit Message.
- Your conventional commit message will be generated and placed directly in your commit message box.
(Optional) Want higher rate limits or custom models?
- Press
Ctrl+Shift+P -> Free AI Commit: Switch Active AI Profile to select Gemini, Groq, Ollama, etc.
- Run
Free AI Commit: Set API Key for Active Profile to enter your personal free tier key (stored securely in VS Code's OS-encrypted keychain).
Provider Profiles & Recommendations
| Rank |
Provider Profile |
Default Model |
Setup Required |
Speed & Limits |
Key Source |
| #1 |
Free Cloud (Default) |
auto |
None (Zero Setup) |
Instant quick-start, daily free pool |
Built-in |
| #2 |
Google Gemini |
gemini-3.5-flash-lite |
Free API Key |
1,500 req/day (30 RPM) |
Google AI Studio |
| #3 |
Groq Cloud |
openai/gpt-oss-120b |
Free API Key |
14,400 req/day (30 RPM, ~300ms latency) |
Groq Console |
| #4 |
Ollama |
qwen2.5-coder:3b |
Local Server |
Unlimited (100% offline, zero data leaves machine) |
Ollama |
| #5 |
OpenRouter Free |
cohere/north-mini-code:free |
Free API Key |
Free community tier |
OpenRouter Keys |
| #6 |
GitHub Models |
gpt-4o-mini |
GitHub PAT |
150 req/day (15 RPM) |
GitHub PAT Tokens |
| — |
DeepSeek |
deepseek-chat |
BYOK |
High reasoning, low cost |
DeepSeek Platform |
| — |
OpenAI |
gpt-4o-mini |
BYOK |
Direct OpenAI API |
OpenAI Platform |
| — |
Anthropic Claude |
claude-3-5-haiku |
BYOK |
Direct Anthropic API |
Anthropic Console |
Recommendation: Keep Free (No Setup Required) for effortless commits right after installing. For heavy day-to-day use with massive diffs, switch to your own free Google Gemini or Groq Cloud key for up to 14,400 requests/day.
Commands
Open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P):
Free AI Commit: Generate AI Commit Message: Generates and populates commit message from staged changes.Free AI Commit: Switch Active AI Profile: Quickly switch between configured providers.Free AI Commit: Set API Key for Active Profile: Securely save or update your API key in the OS keychain.Free AI Commit: Delete API Key for Active Profile: Remove stored credential for active profile from keychain.Free AI Commit: Select Model for Active Profile: Browse and select available models for OpenAI-compatible providers.
Configuration
Open VS Code Settings (Ctrl+, / Cmd+,) and search for Free AI Commit:
| Setting |
Default |
Description |
aiCommitMessage.activeProfile |
"free" |
Active provider preset (free, gemini, groq, github, openrouter, ollama, deepseek, openai, claude). |
aiCommitMessage.language |
"English" |
Output language (supports 19 languages). |
aiCommitMessage.enableGitmoji |
false |
Prefix conventional commit types with Gitmoji symbols (e.g. feat:, fix:). |
aiCommitMessage.customSystemPrompt |
"" |
Optional system prompt override. |
aiCommitMessage.maxDiffCharacters |
60000 |
Safety limit on diff size sent to LLM (minimum 1000). |
aiCommitMessage.diffOverflowStrategy |
"truncate" |
truncate keeps diff head+tail; fail aborts with error. |
aiCommitMessage.temperature |
0.7 |
Sampling temperature (0.0 to 2.0). |
aiCommitMessage.requestTimeoutMs |
120000 |
Request timeout in milliseconds (1s to 600s). |
aiCommitMessage.profiles |
{} |
Custom endpoints and model overrides. |
Custom Endpoints (vLLM, LM Studio, Private Proxies)
Add custom profiles to your settings.json:
{
"aiCommitMessage.profiles": {
"lmstudio": {
"kind": "openai-compatible",
"baseUrl": "http://localhost:1234/v1",
"model": "local-model"
}
}
}
Security & Privacy
- Zero-Setup Quick Start: The default
free profile connects securely over HTTPS to the proxy with server-side prompt isolation, input diff truncation, and zero tracking.
- Encrypted Secret Storage: Personal API keys are stored exclusively in VS Code's OS-encrypted SecretStorage keychain. They are never written to settings JSON files or extension logs.
- Local Isolation with Ollama: When absolute offline privacy is required, select the Ollama profile (
http://localhost:11434/v1). Zero code or metadata leaves your local workstation.
- Prompt Isolation: Staged diffs, file lists, and user notes are strictly delimited and marked as untrusted input to defend against prompt injection.
License
MIT © Christian Gennari
| |
