briefly
PainterLeaf is an open source digital painting application designed to provide artists and creators with efficient and intuitive painting tools. The project supports multi-level drawing, allowing users to customize brushes, color palettes and painting tools to adapt to different creative needs. PainterLeaf also includes image editing functions to help users adjust the color, contrast and brightness of the image. As a lightweight tool, it is committed to providing a concise and powerful creative experience suitable for use by beginners to professionals. Through open source code, community members can participate in development and improvement, so that its functions can be continuously improved and improved.
There are currently too many free models of AI painting, and there are also many site services. Here are three open source projects as examples to create your own free AI painting website.
Painter Leaf
Project address:https://github.com/LeafYeeXYZ/PainterLeaf
Demonstration:https://painter.leafyee.xyz/
1. Functional characteristics
- Prompt-to-image: Supports multiple models, such as Flux.1, StableDiffusion 3.5, etc. Users only need to enter the prompt word to generate the corresponding image.
- Image-to-Prompt: Ability to convert local pictures into prompts. However, before using this feature, you may need to initialize Cloudflare AI’s llama3.2 11B vision model.
- AI Translation: Allow users to use Chinese prompt words in any model to provide convenience for Chinese users.
- Storage function (Store): It saves users ‘creative content to IndexedDB, making it convenient for users to view and manage their works at any time.
2. Deployment method
- Fullstack mode: Environment variables need to be set in the.env file or Vercel. Among them, CF_USER_ID (Cloudflare user ID) and CF_AI_API_KEY (Cloudflare AI API key) are required items, and HF_API_KEY (HuggingFace API key) is optional items. However, since the Vercel Free Plan has a 10-second limit for each request, 504 errors may occur when using the HuggingFace model. Solutions include subscribing to a Vercel paid plan, running the server locally, or adopting a client-server model.
- Client-Server mode: Deploy the server to Cloudflare Workers first, and then set the NEXT_PUBLIC_WORKERS_SERVER environment variable in the.env file or Vercel. When this is set, other environment variables will be ignored.
- Deploy to Vercel: When deploying a project to Vercel, be sure to set environment variables to ensure that the application runs normally.
3. Common problems and solutions
- 429 Error: This is usually due to the request limit of the HuggingFace API being exceeded. The solution is to wait a while, reduce the frequency of requests, or consider a subscription payment plan.
- 504 error: Generally, the request exceeds Vercel’s time limit. You can refer to solutions for Vercel’s time limit, such as switching deployment modes.
4. Development Process
- Cloning a warehouse: using git clone The https://github.com/LeafYeeXYZ/PainterLeaf.git command clones the project warehouse and enters the project directory cd PainterLeaf。
- Install dependencies: You can use the bun install command, or you can choose another package manager you like to install the dependencies you need for your project.
- Local development: Execute the bun run dev command to start the local development environment.
- Build the project: Run the bun run build command to build the project.
Project name: PSHuman
Project function: Image generation character model
Project Information: An open source tool for reconstructing realistic 3D human models from a single image. Ability to generate detailed geometric shapes and a realistic 3D human appearance in one minute, suitable for characters in different postures.
By providing an image of a person wearing clothing, PSHuman can quickly generate a corresponding 3D model.
GitHub：https://github.com/pengHTYX/PSHuman
Oil tubing: