Frequently Asked Questions
Everything you need to know about GIDE
Privacy & Security
Yes! All AI processing happens locally on your machine. GIDE never sends your code or prompts to external servers. The only features that require internet are login authentication and checking for updates.
Absolutely. Your code never leaves your machine. All AI processing is done locally, ensuring complete privacy and security for your sensitive projects.
AI & Models
GIDE ships with Ornith 9B by default, optimized with llama.cpp for efficient local inference. If you want other models, you can download additional ones — Qwen, Gemma, DeepSeek, or any model in GGUF format.
Currently, GIDE uses a bundled model optimized for performance. Custom model support may be available in future releases.
Compatibility
GIDE is compatible with most VS Code extensions available on Open VSX. Some extensions that rely on proprietary Microsoft services may not be available.
GIDE inherits VS Code's language support and can work with virtually any programming language. The AI features work best with popular languages like Python, JavaScript, TypeScript, Java, C++, Go, and Rust.
Getting Started
GIDE requires 16GB RAM and a modern multi-core CPU. Windows 11, macOS 13+ (Apple Silicon M2+), and Linux are supported. To run the bundled Ornith 9B locally, a GPU with 12GB+ VRAM (NVIDIA RTX 3060 12GB or better) is recommended; 8GB cards are limited to smaller downloadable models. Cloud-only mode via BYOK needs no GPU.
Yes! GIDE v1.105.4 added full Linux support. Download as AppImage, .deb, or .rpm from generativeide.com/download. Validated on Ubuntu 24.04 including WSL2 environments.
GIDE bundles a local AI model (Ornith 9B via node-llama-cpp) inside the application. All AI inference runs on your local CPU/GPU — no network request is made to any AI API. The model loads at startup and responds instantly without any internet connection.
GIDE will notify you when updates are available. You can also download the latest version from our website.
Licensing & Pricing
Yes. GIDE offers paid plans, and new users get a 30-day free trial on any paid plan — it runs for 30 days or until you purchase a plan, whichever comes first. You are not charged during the trial period; if you purchase a plan, billing starts immediately and the trial ends. Since AI inference runs locally, there are no per-token API fees on any plan. Check generativeide.com/pricing for current plans.
No. Since all AI processing runs locally on your machine, there are no token costs or usage limits. No developer should ever be charged per AI token.
Yes. GIDE can be used for both personal and commercial projects.
Unlike cloud-based solutions such as GitHub Copilot, Cursor, and Windsurf, GIDE runs entirely on your local machine. This means zero latency, complete privacy, no token costs, and the ability to work without internet. Your code never touches external servers.