Edit photos by describing the change, on your own GPU. A free, open-source alternative to Nano Banana and ChatGPT image editing: photos stay on your computer, no daily limit. One installer for Windows, Mac and Linux sets up the best open model (Qwen-Image 2.1). MCP server for AI agents.
TypeScript ai-photo-editorcomfyui-alternativegenerative-fillimage-editingimage-generationinpaintinglinuxlocal-aimacosmcpmcp-servernano-banana-alternativeofflinephoto-editorprivacyqwen-imageqwen-image-2-1stable-diffusion-cppvulkanwindows
# Kokon
> Kokon is a free, open-source desktop app for editing and creating photos with AI on your own computer, a free, local alternative to Nano Banana and ChatGPT image editing: photos stay on the user's computer and there is no daily limit. Kokon picks and sets up the model for you. You describe the change in plain words ("remove the scooter", "make it a snowy night") and Qwen-Image 2.1 does it locally on your GPU. There is nothing to set up: one installer detects the hardware, downloads and verifies the right model, and picks the resolution and the VRAM/RAM split for every job. Nothing is uploaded.
Key facts (version 0.12.0, November 2026):
- Model: Qwen-Image 2.1 Turbo (official 8-step version, 4-bit GGUF) with Qwen3-VL 8B as the text encoder, run through stable-diffusion.cpp. About 12 GB of models, downloaded once and checked with SHA-256.
- Platforms: Windows 10/11 with an NVIDIA GPU (12 GB VRAM or more, CUDA; 8 GB in beta, about 1-megapixel images). AMD and Intel Arc (12 GB+) through Vulkan, beta. Apple Silicon Macs on macOS 15 or newer (Metal), beta: 16 GB works slowly, 32 GB recommended. Linux x86-64 (.deb for Ubuntu/Debian/Mint, .rpm for Fedora; CUDA or Vulkan), beta.
- Speed: RTX 5070 Ti about 24 s per edit at 1.5K and 36 s for a new image at native 2K; RTX 5070 12 GB about 12 s per 1-megapixel image; M2 Pro 16 GB about 65 s per new 1-megapixel image with Kokon's own Mac engine (about 106 s per edit).
- Features: Prompt+ (a few words in any language become a detailed prompt, written by the Qwen3-VL text encoder already loaded, inside Kokon's own build of stable-diffusion.cpp, so no second model), your own Qwen-Image 2.1 LoRAs (beta, tested on Windows), create from text, new images from reference photos (a person, a product, a style), edit with a sentence, paint an area and Change/Remove/Replace/Add only there, combine several images ("image 1", "image 2"), expand beyond the borders (outpainting), AI upscaling, versions with undo and before/after compare, private mode, works offline after setup.
- For AI agents: built-in MCP server (generate_image, edit_image, inpaint_image, outpaint_image, upscale_image, ...) and a local HTTP API. `npx -y kokon-mcp`; listed in the official MCP Registry as io.github.GabrieleLozupone/kokon.
- License: the app is AGPL-3.0. The Qwen-Image 2.1 weights are under the Qwen Research License, which does not allow commercial use.
- Layers lab (0.12, experimental): click the objects of any image; Kokon cuts each onto its own layer and paints the background behind it; move, hide, edit or expand each layer, blend objects into the scene, export a layered PSD. Also in 0.12: paint with colours (one colour per thing, say what goes there), Kokon's own Mac engine as the default on Apple Silicon, AI marking in every image (IPTC metadata plus an invisible watermark).
- Other devices (0.11): the interface in a browser on a phone, tablet or laptop on the same network, behind a password (off by default); headless `kokon-service` on Linux; Docker image ghcr.io/gabrielelozupone/kokon for NVIDIA GPUs; Tailscale or another VPN for access from outside the home, never an open port.
- Not yet: a Docker image for AMD/Intel (Vulkan); code signing (Windows SmartScreen and macOS Gatekeeper warn on first run).
- Made by Gabriele Lozupone, an AI researcher. Independent project, not affiliated with Alibaba or the Qwen team.
## Pages
- [Home](https://kokon.app/): what Kokon does, examples, requirements, FAQ
- [Free local AI photo editor](https://kokon.app/local-ai-photo-editor/): why edit photos with AI locally, what Kokon can do (remove, relight, text, replace, restyle, expand), requirements, honest limits
- [Offline alternative to Nano Banana](https://kokon.app/offline-alternative-to-nano-banana/): Kokon vs Nano Banana / ChatGPT image editing: what's the same, what's different, when the cloud is still better
- [Simple ComfyUI alternative](https://kokon.app/comfyui-alternative/): what Kokon decides for you (model, files, steps, VRAM), reusing ComfyUI model files, when to stay with ComfyUI
- [Remove objects from photos offline](https://kokon.app/remove-objects-from-photos-offline/): step-by-step object and people removal on your own GPU
- [Split an image into layers](https://kokon.app/image-to-layers/): image to layers and a layered PSD, offline: how it works, the filled background, PSD export, FAQ
- [Uninstall Kokon completely](https://kokon.app/uninstall/): remove the app and every folder it uses, models included, on Windows, Linux and macOS
- [Press kit](https://kokon.app/press/): one-line and short descriptions, key facts, videos, images, logos
- [Run Qwen-Image 2.1 locally](https://kokon.app/qwen-image-2-1-local/): step-by-step setup on Windows, Mac and Linux without Python or ComfyUI, speed, troubleshooting
- [Prompt guide](https://kokon.app/prompt-guide/): how to write prompts for Qwen-Image 2.1, with vague vs clear examples (negative prompts, text in images, edit prompts)
- [Use Kokon from your phone, or run it as a server](https://kokon.app/self-hosted/): browser access from other devices, headless Linux, Docker for NVIDIA, Tailscale
- [Edit photos with AI on your phone, using your own PC's GPU](https://kokon.app/ai-image-editing-from-your-phone/): Settings → Other devices, QR code, password, home-screen app; compared with ComfyUI + --listen + a mobile front end
- [Free local image generation for AI agents (MCP)](https://kokon.app/mcp-image-generation/): use Kokon from Claude Code, Cursor or any MCP client
## Code and downloads
- [GitHub repository](https://github.com/GabrieleLozupone/kokon): source, README, issues
- [Latest release](https://github.com/GabrieleLozupone/kokon/releases/latest): Windows installer and Mac app
- [npm package kokon-mcp](https://www.npmjs.com/package/kokon-mcp): MCP server launcher
- [Demo video](https://www.youtube.com/watch?v=y2NAJkZ4vH0): 85 seconds
- [What's new in 0.11](https://youtu.be/XnC_7P2_cww): Kokon on your phone, 77 seconds
- [What's new in 0.10](https://www.youtube.com/watch?v=qGiGXasJJNI): 42 seconds