You can now create videos, shorts, and voiceovers through the Jellypod API, and videos and shorts through the MCP connector, alongside podcasts.
Turn a training document into a narrated voiceover through the API, or ask your assistant to make a short explaining its key idea. Videos and shorts have their own endpoints and MCP tools, matching the choices in Studio; voiceovers remain an API and Studio product. Both interfaces can discover the available settings, start generation, check progress, and retrieve a download. Failed generations remain failed over MCP; the public API retains its retry endpoints.
File uploads go through the API or the Studio Asset Library. An assistant using MCP can pass a public webpage or YouTube URL, or plain text it already has, as research material, and can also send a finished script directly. Compatible MCP clients also show a product card with status and a preview when the output is ready.
Saved request keys help prevent duplicate generation when a response is lost. Existing podcast integrations can keep their current requests, with optional idempotency keys available for episode generation.
Open Developers for JavaScript, Python, and cURL examples. The product API guide covers the endpoints, and the agent guide explains how to use them through HTTP or MCP.
