GitMind: AI Commit Message Generator for VS Code

GitMind analyzes your Git diff and generates clear, professional commit messages inside VS Code. Choose from 17 built-in AI providers, run locally with Ollama, reuse GitHub Copilot, or connect a custom OpenAI-compatible API with GitMind Pro.

Website: gitmind-pro.com · Documentation: GitMind Handbook · Quick Start · Providers · Native Wiki mirror

Install from the VS Code Marketplace

_{See your provider, model, commit preferences, security, and API status at a glance.}

Highlights

• 18 provider options: OpenAI, Anthropic, NVIDIA NIM, Google Gemini, MiniMax, DeepSeek, xAI Grok, Groq, Perplexity, Z.ai, Mistral, Cohere, Hugging Face, Together AI, OpenRouter, Ollama, GitHub Copilot, and Custom API.
• Searchable, dynamic model selection: Load current models from supported provider APIs and quickly filter large model catalogs.
• Professional commit standards: Conventional Commits, Angular, Semantic Release, Gitmoji, Linux Kernel, jQuery, Ember.js, and more.
• 3-Candidate Choice QuickPick with Live Multiline Preview: Compare 3 distinct AI candidate archetypes (Concise, Detailed, Intent) with full un-truncated multiline live preview as you navigate choices.
• Guided Reviewed Commit Workspace: Interactive 3-step dashboard with branch badges, file line change stats (+24/-10), noise classification, bulk context controls, and Commit Health rating meter (0–100).
• Pre-Commit Code Review Panel: Real-time pre-commit code review findings with color-coded severity callouts (ERROR, WARNING, INFO).
• Flexible Git workflow: Generate from staged changes or enable Capture All Changes to include unstaged and untracked files.
• Local and key-free options: Use Ollama locally or an existing GitHub Copilot subscription.
• Large diff support: Token-aware processing keeps generation useful on substantial changes.

Explore GitMind

Models And Automatic Recovery Configure NVIDIA NIM and other providers, load searchable models, retry eligible temporary failures, and select a provider-scoped fallback model.	Professional Commit Styles Choose Conventional Commits, Angular, Gitmoji, Semantic Release, Ember.js, and other structured formats.
Security And Model Control Encrypt API keys, process large diffs, and tune advanced generation parameters when you need precise control.	Match Your Team Generate in your target language, control message length, and learn from existing commit history.
Useful Free Features Control verbose messages, capture all changes, custom context, and generation diagnostics.	AI Changelog Generation Generate professional changelogs from Git history with version grouping and configurable commit ranges.

Quick Start

1. Install GitMind from the VS Code Marketplace.
2. Open GitMind Settings and select an AI provider.
3. Add the provider API key when required, then load or search for a model.
4. Stage changes, or enable Capture All Changes.
5. Click the GitMind icon in Source Control or run Generate GitMind Commit Message.

ext install ShahabBahreiniJangjoo.ai-commit-assistant

Supported Providers

Provider catalogs change frequently. GitMind loads current model lists where the provider supports discovery and falls back to known compatible models when necessary.

Free And Pro

Automatic Recovery

GitMind Pro can recover from selected generation failures without creating retry loops:

• Retries once for network failures, timeouts, and temporary provider failures.
• Can switch once to a configured fallback model for model-specific limits and ordinary HTTP 429 rate limits.
• Does not retry invalid API keys, billing/account quota exhaustion, permission errors, malformed requests, or content-policy failures.
• Shows a clear notification explaining the failure, recovery action, and final result.

The fallback model picker is searchable and scoped to the currently selected provider.

NVIDIA NIM

GitMind supports NVIDIA hosted NIM through its OpenAI-compatible LLM API.

1. Create an API key at NVIDIA Build.
2. Select NVIDIA in Model Settings.
3. Add the key and load the available hosted NIM models.
4. Search for a model and save your settings.

See the NVIDIA NIM LLM API reference for provider details.

GitMind Pro Activation

Open GitMind Settings > Pro and paste the license key from your purchase email.

GitMind Pro is a one-time lifetime purchase available at gitmind-pro.com/pricing. You can manage your licenses and active devices at gitmind-pro.com/portal. Activation, deactivation, and current Pro status are also available directly in the settings panel.

If your key does not work, or you cannot find it, open an issue at GitMind-Community Issues and we will sort it out — you will not lose access while we do.

Privacy And Security

• GitMind sends the selected Git diff and prompt to the provider you configure.
• Ollama can keep generation local.
• GitMind Pro can store provider keys in VS Code SecretStorage.
• GitMind does not collect or transmit product telemetry.
• Pro users can capture a bounded, local support report containing only allowlisted operational metadata. It excludes source, diffs, prompts, paths, URLs, credentials, request/response bodies, raw errors, email, and license/customer data.

Requirements

• VS Code 1.101.0 or newer
• Git repository
• API key for the selected cloud provider, unless using Ollama or GitHub Copilot

Support

• Official Website
• Read the GitMind Handbook
• For serious debugging, Pro users can open GitMind Settings > Pro > Support Report, start a session, reproduce the issue, stop and review it, then save and manually attach the JSON report.
• Use the native GitHub Wiki mirror
• Report an issue
• Sponsor development