One MCP server gives your assistant all 450+ models. Claude and ChatGPT connect by signing in; editors and agents connect with a key.
Claude
Make images and videos inside a Claude chat on the web, desktop or mobile.
Customize > Connectors > Add custom connector, paste the sign-in address from our MCP page, then Connect and Allow.
ChatGPT
Generate media from ChatGPT with a custom connector.
Turn on Developer mode under Settings > Apps, create a connector with the sign-in address and OAuth, then sign in and Allow.
Open WebUI
A GenMagic tool that makes images and videos with sound in any Open WebUI chat, with the model picked per request.
Open Workspace > Tools > +, paste the tool's code and save, then set your own API key in the tool's settings (Valves). For images only, Open WebUI's own image settings work too: Admin Panel > Settings > Images, engine OpenAI, base URL https://genmagic.co/api/v1, your GenMagic key and a model such as google/gemini-3-pro-image.
NextChat
Chat on GenMagic's text models and images from its DALL-E 3 option, which runs on GPT Image 1 Mini.
In NextChat's settings turn on a custom endpoint for OpenAI, set the endpoint to https://genmagic.co/api (NextChat adds /v1 itself) and the key to your GenMagic key; OpenAI model names such as gpt-4o-mini resolve.
LobeHub
Chats on GenMagic's text models under OpenAI's names, and pictures on its Image page from the OpenAI image models it lists (GPT Image 1, 1 Mini, 2 and 2.5).
In LobeHub's settings open Provider, then OpenAI, and set API Key to your GenMagic key and API Proxy URL to https://genmagic.co/api/v1.
ONLYOFFICE
Pictures and text in ONLYOFFICE's document, spreadsheet and presentation editors through its AI plugin, on GenMagic's models.
In the AI plugin's settings choose Edit AI models and add a model with Provider OpenAI, URL https://genmagic.co/api/v1 and your GenMagic key. Pick a GPT Image model such as openai/gpt-image-1-mini under Model and use it for images; add a second model the same way with any text model for text.
Moodle
Teachers make pictures in Moodle's text editor, and its AI actions write and summarise course text, on GenMagic's models (Moodle 4.5 and later).
In Site administration, AI, AI providers, create an OpenAI API provider with your GenMagic key. In its Actions, set Generate image's API endpoint to https://genmagic.co/api/v1/images/generations and its AI model to dall-e-3 or a model id such as google/gemini-2.5-flash-image, and the text actions' API endpoint to https://genmagic.co/api/v1/chat/completions (gpt-4o works as named). Then turn on Generate AI images in text editor under AI placements.
Chamilo
Teachers in Chamilo 2 make pictures and short videos with its AI image and video generators, and its text helpers run, on GenMagic's models.
In Administration, Settings, AI helpers, add GenMagic to the ai_providers JSON as openai with api_key your GenMagic key and these urls: text https://genmagic.co/api/v1/chat/completions; image https://genmagic.co/api/v1/images/generations with a model such as google/gemini-2.5-flash-image; video https://genmagic.co/api/v1/videos with status_url https://genmagic.co/api/v1/videos/{id} and content_url https://genmagic.co/api/v1/videos/{id}/content and model sora-2, which GenMagic runs on Veo 3.1 Fast. Then turn on the image and video generators.
Chatbox
Chat on GenMagic's text models and make pictures in Chatbox's Image Creator with GenMagic's image models.
In Chatbox go to Settings, Model Provider and Add a provider with API Mode OpenAI API Compatible, API Host https://genmagic.co/api/v1 and your GenMagic key; Fetch loads the chat models, and a New model with Model Type Image (for example google/gemini-2.5-flash-image) appears in Image Creator.
LibreChat
LibreChat's own image tool makes and edits images in any chat with GenMagic's image models.
In LibreChat's .env set IMAGE_GEN_OAI_BASEURL to https://genmagic.co/api/v1 and IMAGE_GEN_OAI_API_KEY to your GenMagic key (optionally IMAGE_GEN_OAI_MODEL to a model id such as google/gemini-3-pro-image), then turn on the OpenAI Image Tools.
Agent Skills
Eight skills that teach your coding agent to make images, video, speech and music with GenMagic, including faceless shorts, UGC ads, n8n workflows and moving Sora 2 code to a working model. Works in Claude Code, Codex, Cursor, Copilot, Windsurf, Gemini, Cline, OpenCode and Hermes.
Run npx skills add https://genmagic.co/ and pick the skills you want; the agent reads your key from GENMAGIC_API_KEY.
Codex and Claude Code plugin
One plugin that connects GenMagic's MCP server with sign-in and adds a skill telling the agent when to make media and what it costs.
Codex: codex plugin marketplace add https://codeberg.org/genmagic/genmagic-plugin.git, then codex plugin add genmagic@genmagic-media. Claude Code: claude plugin marketplace add with the same address, then claude plugin install genmagic@genmagic-media.
Claude Code, Cursor and VS Code
Let your coding agent make the images, icons and clips a project needs.
One-click install links for Cursor and VS Code, or one command for Claude Code, all with your API key.
Context7
GenMagic's API reference is indexed on Context7, the docs index coding agents read, so an agent asked to use GenMagic writes calls that work.
With Context7 connected to your agent, ask it to use context7 for GenMagic, or point it at the library /openapi/genmagic_co_openapi_json.