MCP server and command-line tool

Use chat, image and video models from an AI agent through the hosted MCP server, or from a terminal with the apistore command-line tool.

MCP

Use the models from an AI agent through our hosted MCP server (Streamable HTTP). Add the URL below to Claude Code, Cursor, Windsurf, VS Code or any MCP client, with your API key in the Authorization header. Calls are billed to your balance at the same prices as the API.

URL: https://api.apistore.cloud/mcp
Authorization: Bearer sk-...
# Claude Code
claude mcp add --transport http apistore https://api.apistore.cloud/mcp \
  --header "Authorization: Bearer sk-..."

# Cursor (~/.cursor/mcp.json), Windsurf, VS Code and other MCP clients
{
  "mcpServers": {
    "apistore": {
      "url": "https://api.apistore.cloud/mcp",
      "headers": { "Authorization": "Bearer sk-..." }
    }
  }
}

Tools: list_models (models and your prices), chat (send a prompt or messages to a model and get its answer), create_image (make images and get their links), create_video (start a video and get its id) and get_video (its status, then the link to the file when it is done).

Setup guides: Cursor · Cline · Claude Code

CLI

The apistore command runs chat, image and video models from a terminal or a shell script: text in, answer out, files saved where you say. It needs Node 20 or newer.

npm install -g @apistorecloud/apistore
apistore login --key sk-...

apistore chat -m deepseek-v4-1-flash "Hello!"
apistore chat "Summarize this" < notes.md
apistore image "a red bicycle on a white wall" -o bike.png
apistore video "waves at sunset, slow pan" --resolution 720p --seconds 5

Every command takes --json for the raw API response, and the key can also come from APISTORE_API_KEY for scripts and CI. Run apistore --help for all options.