The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS.
description: >- Ask for an image or a video and dedicated media agents generate it, with image editing and animation of a reference image, saved in your Files folder. icon: clapperboard
Image and video generation
OpenHuman can make media, not just read it. Ask for "an image of…", "edit this screenshot to…" or "animate this photo into a short clip", and a dedicated media sub-agent takes over. You need no plugin, no API key and no separate billing.
What it can do
- Image generation and editing. Text-to-image and image editing through hosted GMI models (Seedream for generation, SeedEdit for edits).
- Video generation. Text-to-video, or animate a reference image into a clip (Seedance or Veo). Video is asynchronous. The agent starts the render and collects the clip when it is done.
- Model discovery. The agent can list the media models available now and pick the right one.
How it works
The media_generation domain (crates/openhuman-core/src/media/generation/) gives the agent three tools: generate image, generate video and list models. They use the OpenHuman backend's media-generation provider. The backend owns the provider keys, billing and rate limiting, and your subscription covers it like any other model call.
Images poll every 4 seconds for up to 180 seconds; video polls every 5 seconds for up to 600 seconds. You and the agent see live progress instead of a hung call. Finished files are filed in the Files folder and returned as local file paths, ready to attach, post or edit further.
Privacy
Prompts and reference media go to the OpenHuman backend and on to the hosted media provider. The in-app capability catalog discloses this (intelligence.image_generation and intelligence.video_generation, both Beta).
Privacy mode local-only enforcement covers inference providers only. The media tools still call the backend, so avoid them if you need strict no-egress today.
See also
- Image tools: attachments, image metadata and the vision model slot.
- Native tools: the full toolbelt.
- Billing, cost and usage: how media jobs are metered.