VIDU Q4 · PREVIEW
Vidu Q4 Preview: Image-to-Video Guide & Credits
Create 3–16 second videos from one image with Vidu Q4 Preview on VidVid. Compare 540p–4K credits, audio settings and prompt workflows.
Start at 70 credits for 3 seconds at 540p. A 5-second 720p clip costs 245 credits. Preview capabilities and provider pricing can change.
Create with this modelChoose the right Q4 workflow
Upload one finished frame, then describe the movement that should happen next. Output framing follows the source image. Enable prompt expansion when you want the provider to elaborate your instruction; disable it when comparing precise prompt variants.
- Choose the model and upload the required source images.
- Describe one main action, camera movement and desired sound. Select duration, resolution and audio.
- Check the credits on Generate, submit, then review the completed task in history.
Vidu Q4 Preview credit prices
Both Q4 workflows use the same duration and resolution pricing. Audio adds no credits. The table shows VidVid credits per generation, not WaveSpeed subscription prices.
| Duration | 540P | 720P | 1080P | 2K | 4K |
|---|---|---|---|---|---|
| 3s | 70 | 147 | 186 | 294 | 602 |
| 5s | 116 | 245 | 309 | 489 | 1003 |
| 10s | 232 | 489 | 618 | 978 | 2006 |
| 16s | 371 | 782 | 988 | 1564 | 3210 |
All integer durations from 3–16 seconds are available. The generator calculates the exact charge before submission. View credit plans
Plan a draft-to-final budget
Two 5-second 540p drafts plus one 5-second 1080p render cost 541 credits (116 + 116 + 309). Each new generation is charged separately and may produce a different result; increasing resolution is not an upscale of the previous clip.
Prompt starting points
These are illustrative prompts, not tested output examples. Start with a short lower-resolution draft and inspect shape, motion and sound before a larger render.
Keep the product silhouette from the source image. Over five seconds, slowly orbit the camera a quarter turn; soft light moves across the packaging. Quiet room ambience, no dialogue.
For a source-image shot, describe what changes over time rather than repeating the still image. Keep important subjects inside the source frame; there is no separate aspect-ratio control in this mode.
The multi-image example requires Reference mode. VidVid currently exposes image references and generated audio; external audio reference uploads are not available in this interface.
Before you retry a shot
If motion is too busy, ask for one action and one camera move. Check the draft for product shape, hands, faces and audible dialogue before paying for a longer or higher-resolution generation.
Frequently asked questions
Can I generate from text without an image?
Choose Vidu Q4 Preview Reference in text-to-video mode for prompt-only generation. The standard Q4 Preview image-to-video model requires one image.
How many reference images can I upload?
Standard image-to-video takes one image. Reference image-to-video accepts up to 12 images in VidVid, following the WaveSpeed API documentation.
Does generated audio cost extra?
WaveSpeed documents the same rate with audio enabled or disabled. VidVid exposes an audio switch and charges the same credits for both.
Which durations and resolutions are supported?
Integer durations from 3 to 16 seconds; 540p, 720p, 1080p, 2K and 4K. Standard image-to-video follows image framing; Reference supports five selectable aspect ratios.
What happens if submission fails?
Credits are reserved when a task is submitted. If submission fails before the provider accepts it, VidVid refunds that task once. Accepted tasks continue through the generation history and result workflow.
Sources and related models
API specifications and provider prices checked October 9, 2026. No quality ranking or real-generation benchmark is claimed on this page.