🤗 HF Studio

Hugging Face token (hf_...) Chat model (use a vision model to send images)

Image model

Video model

Extra system prompt (optional)

Max tokens

Temperature

Let the AI generate images/videos in chat Remember token on this device

💧 Watermark

Enable (photos and videos) Text Logo image (PNG with transparency works best)

Logo size, % of width

Text size, % of width

Opacity, %

Photo position Show on video (teleports between corners)

Video: stays, sec

Video: fade in/out, sec

On video, the Download button renders an MP4 with the watermark baked in (sound is kept, takes a few seconds). Needs Chrome / Edge. “Original” saves the clean file.

Ask anything, attach images or files with 📎.
Say “draw a cat astronaut” or use /image and /video commands.
Enter = send · Shift+Enter = new line

Prompt Resolution Reference image (optional, image → image)

If set, the image model must support image-to-image editing.

Advanced: extra parameters (JSON)

Prompt Mode Start image (required for image-to-video models like MiniMax-H3)

Resolution

Aspect ratio

Seconds

Blank / “Model default” = not sent. Supported values depend on the model and provider.

Advanced: extra parameters (JSON)