Lightricks
LTX AI Video Generator
Lightricks’ video model with sound: LTX-2.3 Pro and low-cost LTX-2.5 Fast.
AI video model · ShengShu
Vidu is ShengShu’s video model, and Q3 generates sound with the picture. Make Vidu Q3 Pro and Q3 Turbo clips from a prompt or an image inside AutoFeed, vertical, horizontal or square.
Plans from $19/mo · cancel anytime
Vidu is the video generation model family from ShengShu Technology. The Q3 generation generates audio together with the video, including dialogue and sound effects.
AutoFeed offers Vidu Q3 Pro for the highest quality and Vidu Q3 Turbo for fewer credits per clip. Both take a text prompt or a starting image and render 5 or 10 second clips at 720p or 1080p in vertical, horizontal or square format.
Use it for
Credits per clip for every variant, resolution and length, and how many clips a month that is on each plan: Hobby $19/month = 15,000 credits, Pro $59/month = 60,000 credits, Ultimate $249/month = 180,000 credits. Credits are shared with everything else you make.
| Feature | Input | Credits | Clips on Hobby | Clips on Pro | Clips on Ultimate |
|---|---|---|---|---|---|
| Vidu Q3 Pro, 720p, 5s | Text or image | 3,000 | 5 | 20 | 60 |
| Vidu Q3 Pro, 720p, 10s | Text or image | 6,000 | 2 | 10 | 30 |
| Vidu Q3 Pro, 1080p, 5s | Text or image | 3,200 | 4 | 18 | 56 |
| Vidu Q3 Pro, 1080p, 10s | Text or image | 6,400 | 2 | 9 | 28 |
| Vidu Q3 Turbo, 720p, 5s | Text or image | 1,200 | 12 | 50 | 150 |
| Vidu Q3 Turbo, 720p, 10s | Text or image | 2,400 | 6 | 25 | 75 |
| Vidu Q3 Turbo, 1080p, 5s | Text or image | 1,600 | 9 | 37 | 112 |
| Vidu Q3 Turbo, 1080p, 10s | Text or image | 3,200 | 4 | 18 | 56 |
The AI video editor, step by step.
Vidu is developed by ShengShu Technology. AutoFeed runs it through Replicate, so you do not need a Vidu account.
Both take text or an image and make 5 or 10 second clips with sound at 720p or 1080p. Q3 Pro is the higher quality tier; Q3 Turbo costs fewer credits per clip.
From 1,200 credits (a 5 second Vidu Q3 Turbo clip at 720p) to 6,400 credits (a 10 second Vidu Q3 Pro clip at 1080p). The table on this page lists every combination of variant, resolution and length.
Hobby ($19/month, 15,000 credits): 2 to 12 clips; Pro ($59/month, 60,000 credits): 9 to 50 clips; Ultimate ($249/month, 180,000 credits): 28 to 150 clips. Credits are shared across everything you make on AutoFeed, so clips on other models or faceless videos draw on the same balance.
The cheapest Vidu clip costs 1,200 credits, more than the 1,000 free credits a new account receives, so Vidu needs a plan. The free credits do cover a clip on a cheaper model such as Wan 2.2.
No. AutoFeed adds no watermark or logo. Each clip is delivered as an MP4 at 1080x1920 for 9:16, 1920x1080 for 16:9 or 1280x1280 for 1:1, ready to download or post.
Both. Vidu on AutoFeed accepts a text prompt on its own, or a starting image plus a prompt describing the motion.
Yes. A single clip can be posted to a connected TikTok or YouTube account as soon as it is ready, or scheduled for a date and time. To keep a channel going, create an AI Video series instead: pick Vidu, write one or more prompts and choose the days and times, and AutoFeed generates and posts a new clip at every slot. Your prompts are used word for word, and each post costs the clip price shown before you save.
Every model runs on the same credits, in the same editor.
Lightricks
Lightricks’ video model with sound: LTX-2.3 Pro and low-cost LTX-2.5 Fast.
Google’s video model with generated sound, vertical or horizontal.
Kuaishou
Kuaishou’s video model: Kling 3.0, 2.6 with sound and 2.5 Turbo Pro.
ByteDance
ByteDance’s video model: Seedance 2.0 and 2.0 Fast with sound, plus three Seedance 1 tiers.
MiniMax
MiniMax’s video model: Hailuo 2.3 and 02, each with a low-cost Fast variant.
PixVerse
Video model for anime characters and action, in every aspect ratio.
