[](https://bolt.diy)
Sign in to like and favorite skills
Welcome to bolt.diy, the official open source version of Bolt.new, which allows you to choose the LLM that you use for each prompt! Currently, you can use OpenAI, Anthropic, Ollama, OpenRouter, Gemini, LMStudio, Mistral, xAI, HuggingFace, DeepSeek, or Groq models - and it is easily extended to use any other model supported by the Vercel AI SDK! See the instructions below for running this locally and extending it to include more models.
Check the bolt.diy Docs for more offical installation instructions and more informations.
Also this pinned post in our community has a bunch of incredible resources for running and deploying bolt.diy yourself!
We have also launched an experimental agent called the "bolt.diy Expert" that can answer common questions about bolt.diy. Find it here on the oTTomator Live Agent Studio.
bolt.diy was originally started by Cole Medin but has quickly grown into a massive community effort to build the BEST open source AI coding assistant!
Join the bolt.diy community here, in the oTTomator Think Tank!
Bolt.diy is a community effort! Still, the core team of contributors aims at organizing the project in way that allows you to understand where the current areas of focus are.
If you want to know what we are working on, what we are planning to work on, or if you want to contribute to the project, please check the project management guide to get started easily.
app\lib\.server\llm\prompts.tsIf you're new to installing software from GitHub, don't worry! If you encounter any issues, feel free to submit an "issue" using the provided links or improve this documentation by forking the repository, editing the instructions, and submitting a pull request. The following instruction will help you get the stable branch up and running on your local machine in no time.
Let's get you up and running with the stable version of Bolt.DIY!
Before you begin, you'll need to install two important pieces of software:
Node.js is required to run the application.
Windows + RNode.jsecho $PATH
/usr/local/binYou have two options for running Bolt.DIY: directly on your machine or using Docker.
Install Package Manager (pnpm):
npm install -g pnpm
Install Project Dependencies:
pnpm install
Start the Application:
pnpm run dev
This option requires some familiarity with Docker but provides a more isolated environment.
Build the Docker Image:
# Using npm script: npm run dockerbuild # OR using direct Docker command: docker build . --target bolt-ai-development
Run the Container:
docker compose --profile development up
Setting up your API keys in Bolt.DIY is straightforward:
For providers that support custom base URLs (such as Ollama or LM Studio), follow these steps:
Click the settings icon in the sidebar to open the settings menu
Navigate to the "Providers" tab
Search for your provider using the search bar
Enter your custom base URL in the designated field
Note: Custom base URLs are particularly useful when running local instances of AI models or using custom API endpoints.
This method is recommended for developers who want to:
Clone the Repository:
git clone -b stable https://github.com/stackblitz-labs/bolt.diy.git
Navigate to Project Directory:
cd bolt.diy
Install Dependencies:
pnpm install
Start the Development Server:
pnpm run dev
(OPTIONAL) Switch to the Main Branch if you want to use pre-release/testbranch:
git checkout main pnpm install pnpm run dev
Hint: Be aware that this can have beta-features and more likely got bugs than the stable release
Open the WebUI to test (Default: http://localhost:5173)
- Beginngers:
- Try to use a sophisticated Provider/Model like Anthropic with Claude Sonnet 3.x Models to get best results
- Explanation: The System Prompt currently implemented in bolt.diy cant cover the best performance for all providers and models out there. So it works better with some models, then other, even if the models itself are perfect for >programming
- Future: Planned is a Plugin/Extentions-Library so there can be different System Prompts for different Models, which will help to get better results
To get the latest changes from the repository:
Save Your Local Changes (if any):
git stash
Pull Latest Updates:
git pull
Update Dependencies:
pnpm install
Restore Your Local Changes (if any):
git stash pop
If you encounter issues:
Clean Installation:
# Remove node modules and lock files rm -rf node_modules pnpm-lock.yaml # Clear pnpm cache pnpm store prune # Reinstall dependencies pnpm install
Reset Local Changes:
# Discard all local changes git reset --hard origin/main
Remember to always commit your local changes or stash them before pulling updates to avoid conflicts.
pnpm run devpnpm run buildpnpm run startpnpm run previewpnpm testpnpm run typecheckpnpm run typegenpnpm run deploypnpm run lint:fixWe welcome contributions! Check out our Contributing Guide to get started.
Explore upcoming features and priorities on our Roadmap.
For answers to common questions, issues, and to see a list of recommended models, visit our FAQ Page.
Who needs a commercial WebContainer API license?
bolt.diy source code is distributed as MIT, but it uses WebContainers API that requires licensing for production usage in a commercial, for-profit setting. (Prototypes or POCs do not require a commercial license.) If you're using the API to meet the needs of your customers, prospective customers, and/or employees, you need a license to ensure compliance with our Terms of Service. Usage of the API in violation of these terms may result in your access being revoked.