developers

An AI avatar video API that ends in a posted video

Pamba's REST API creates an AI avatar, turns a script into an avatar video, renders the final cut and can schedule it to TikTok and Instagram, all with one workspace API key. Generation spends credits from your plan. The same work is available to agents through Pamba's MCP server.

How does it work?

  1. 1. Make or pick an avatar

    GET /avatars lists your avatars. POST /avatars/wizards starts a new one: describe the person, pick a look and a voice.

  2. 2. Create projects from scripts

    POST /projects creates one or many projects for an avatar, as reviewable drafts or straight to a finished video.

  3. 3. Generate and render

    POST /projects/{projectId}/generate-clip-videos-async generates the clips, then POST /projects/{projectId}/render renders the video. GET /projects/{projectId}/render-jobs/latest reports progress.

  4. 4. Or copy a format

    Replication turns a reference TikTok into a new video with your avatar: POST /projects/{projectId}/replication/generate.

  5. 5. Post it

    Schedule the result with POST /projects/{projectId}/schedule (see the posting API page).

What should you know first?

What it makes
Talking-avatar videos ("yappers"), replicated formats and photo slideshows
Cost
Credits from your plan: Free 300 credits, Hobby $16 a month, Pro $100 a month (see pricing)
Auth
Workspace API key in the X-API-Key header
Agents
Every route has an MCP tool at https://api.pamba.app/mcp

Questions

Do I have to post through Pamba?

No. Generating and rendering do not require scheduling anything. Posting through Pamba's accounts is a separate, optional step.

How much does a video cost?

It depends on length and model; GET /usage/pricing returns the live credit rates and the pricing page shows plan credit totals.

Can I use my own face or voice?

Avatars are described and generated. Do not create an avatar of a real person without their consent; the Acceptable Use Policy covers this.

read next