RunningHub Video Generator
Generate videos from images via RunningHub API.
Generators
Usage
- Enter your API Key
- Upload image(s) and audio (if supported)
- Configure prompt, LoRA, aspect ratio (if available)
- Click “Generate Video”
- Preview → download (result URL valid for 24 hours)
Video Edit (6 Images) Features
- 1 video slot (required) — source video for the edit
- 6 image slots — Picture 1 (required) + Picture 2–6 (optional, workflow defaults used if not selected)
- Prompt tags — reference the video as
<Video 1> and images as <Picture 1>…<Picture 6>
- Default prompt — head swap template pre-filled
- 1 LoRA slot with strength
- Instance type selector — default (24G) / plus (48G) / ultra (84G) VRAM
- Direct video output — result is a ready video file
Video Edit (1 Image) Features
- 1 video slot (required) — source video for the edit
- 1 image slot (required) — reference image
- Prompt tags — reference the video as
<Video 1> and the image as <Picture 1>
- Default prompt — head swap template pre-filled
- 1 LoRA slot with strength
- Direct video output — result is a ready video file
MiniMax H3 Ref2VA (6 Images + Audio) Features
- 6 image slots — Picture 1 (required) + Picture 2–6 (optional, workflow defaults used if not selected)
- 1 audio slot (optional) — audio or mp4
- Prompt tags — reference images as
<Picture 1>…<Picture 6> and audio as <Audio 1> in the prompt
- Default prompt — subject definitions template pre-filled
- 3 LoRA slots with individual strength
- 1 MP output — resolution follows Picture 1, no aspect ratio selector
MiniMax H3 Ref2VA (1 Image + Audio) Features
- 1 image slot (required) — resolution follows Picture 1
- 1 audio slot (optional) — audio or mp4
- Prompt tags — reference the image as
<Picture 1> and audio as <Audio 1> in the prompt
- Default prompt — subject definitions template pre-filled
- 3 LoRA slots with individual strength
- 1 MP output — fixed resolution, no aspect ratio selector
MiniMax H3 Ref2VA (6 Images, Portrait 9:16) Features
- 6 image slots — Picture 1 (required) + Picture 2–6 (optional, workflow defaults used if not selected)
- 1 audio slot (optional) — audio or mp4
- Prompt tags — reference images as
<Picture 1>…<Picture 6> and audio as <Audio 1> in the prompt
- Default prompt — subject definitions template pre-filled
- 3 LoRA slots with individual strength
- 1 MP output — fixed resolution, no aspect ratio selector
- Portrait 9:16 aspect ratio
MiniMax H3 (1 Image) Features
- 1 image slot (required)
- 3 LoRA slots with individual strength
- Aspect ratio selector
MiniMax H3 Ref2VA Features
- 5 image slots — Picture 0 (required) + Picture 1–4 (optional, workflow defaults used if not selected)
- 1 audio slot (optional) — audio or mp4
- Prompt tags — reference images as
<Picture 0>…<Picture 4> and audio as <Audio 0> in the prompt
- 3 LoRA slots with individual strength
- Aspect ratio selector
MiniMax H3 Features
- 3 LoRA slots (LoRA 1/2/3) with individual strength
- Aspect ratio selector (1:1, 2:3, 3:2, 3:4, 4:3, 9:16, 16:9, 21:9)
- Audio support — auto-detects
ComfyUI_*.mp3 in results, FPS calculated from audio duration
- Frame decoding in browser — decodes ~360 frames with watermark removal, exports WebM with audio
- 24 FPS default (vs 16 FPS in Wan2.2 generators)
Share Settings
Click “Share Settings” to copy a URL with your current prompt, LoRA, and strength values. Opening the link auto-fills the fields. API Key is not included in the URL.
Browsers
- Chrome — Full support
- Firefox — Full support (VP8/VP9 auto-detect)
- Safari — Limited