Vidu Q4 Preview — What It Is & How to Try It
On October 7, 2026, ShengShu Technology shipped the first public preview of Vidu Q4, its next-generation flagship video model — and it went live on fal's API the same day. If you follow AI video, that combination (new flagship + instant API access) is the interesting part: you don't have to wait for a waitlist to test it. Here's what Q4 actually does, what it costs, how it stacks against Kling 2.6 and Veo 3.1, and the fastest way to run your first clip.
What Vidu Q4 Preview does
Q4's headline theme is coordination: facial expression, emotion, body movement and generated audio stay locked together, from a quiet glance to full-scale action. The practical specs:
- 16-second clips — the longest of any model on our lineup (Kling tops out at 10s, Veo at 8s)
- Native audio — dialogue and sound effects generated with the video, at no extra price
- Up to 4K — 540p / 720p / 1080p / 2K / 4K on one model
- Multi-image referencing — up to 12 reference images for character and object consistency (the model also accepts voice reference clips; img2.video deliberately does not expose voice-clone inputs)
Vidu Q4 API pricing (fal)
Billed per second; a 30% promo runs through November 30, 2026. Audio doesn't change the price:
| Resolution | Promo (to Nov 30) | Standard | 5s clip @ standard |
|---|---|---|---|
| 720p | $0.0665/s | $0.095/s | $0.475 |
| 1080p | $0.084/s | $0.12/s | $0.60 |
| 4K | $0.273/s | $0.39/s | $1.95 |
On img2.video, credits make this simple: a 5s 720p Vidu Q4 clip costs 30 credits — and new accounts get 30 free.
Vidu Q4 vs Kling 2.6 vs Veo 3.1
| Model | Max length | Audio | Max res | From |
|---|---|---|---|---|
| Vidu Q4 | 16s | Native, free | 4K | 30 credits |
| Kling 2.6 Pro | 10s | Native, 2× per-sec | 720p | 20 credits |
| Veo 3.1 Fast | 8s | Native, +50% | 4K | 30 credits |
Quick take: Q4 for long scenes and free audio, Kling for cinematic motion per credit, Veo for maximum polish. Full breakdown in our model comparison.
How to try Vidu Q4 free
- Open the Vidu Q4 generator on img2.video.
- Upload a photo (front-facing subject, face large in frame works best).
- Describe the motion — e.g. "the subject turns and laughs as warm light sweeps across the room, ambient sound on".
- Pick 5s / 720p (your 30 free credits cover it exactly), enable audio, generate, download the clean MP4.
It's a preview model — outputs may shift week to week as ShengShu tunes it. On our side pricing stays fixed and failures auto-refund, so experimenting costs you nothing.
FAQ
What is the Vidu Q4 preview?
Vidu Q4 Preview is the public preview of ShengShu Technology’s next-generation flagship AI video model, released October 7, 2026. It generates video from images with coordinated facial expression, emotion, body motion and native audio — clips up to 16 seconds at up to 4K.
How much does the Vidu Q4 API cost?
On fal’s API, Vidu Q4 is billed per second. Standard rates: 540p $0.045/s, 720p $0.095/s, 1080p $0.12/s, 2K $0.19/s, 4K $0.39/s — with a 30% promo discount through November 30, 2026. Generating audio does not change the price.
Is Vidu Q4 free to try?
You can try it on img2.video with 30 free credits on sign-up — enough for a full 5-second 720p clip, no card required.
Vidu Q4 vs Kling 2.6 vs Veo 3.1 — which is better?
Vidu Q4 wins on clip length (16s) and free native audio; Kling 2.6 is the value pick for cinematic motion; Veo 3.1 leads on visual polish in 4K. All three run side by side on img2.video.
Can Vidu Q4 clone voices?
The underlying model supports voice reference clips, but img2.video does not expose voice-clone inputs — we offer image + prompt generation with an optional native-audio toggle only, consistent with our policy against impersonation tools.
Is a preview model safe to build on?
Preview means the model may change as ShengShu iterates. On img2.video, pricing stays fixed and failed generations are refunded automatically — so the risk of trying it is effectively zero.
Sources: Vidu Q4 official page · Vidu Q4 API & pricing (fal.ai) · Launch announcement (PR Newswire). Last reviewed October 2026.