# VEED Fabric 1.0

> VEED Fabric 1.0 is VEED's image-to-video model that turns one still image and an audio track into a talking video, with lips, head and expression following the speech. VEED launched it in September 2025. It works with photos, illustrations, animals and anime characters. On Kenerate AI it runs in Avatar Talk for up to 5 minutes per video, at 21 credits a second at 480p or 42 at 720p, billed on the length of the speech.

URL: https://kenerateai.com/models/veed/fabric-1-0
Publisher: Kenerate AI (https://kenerateai.com)
Last updated: 20 October 2026

## Key facts
- Official name: VEED Fabric 1.0
- API ID: veed/fabric-1.0
- Maker: VEED
- Released: September 2025
- Inputs: 1 image + 1 speech track
- Resolution: 480p or 720p
- Length: 3 s to 5 min of speech
- Works on: Photos, illustrations, animals, anime
- On Kenerate: 21 cr/s (480p) · 42 cr/s (720p)
- Billing: Per second of speech, min 3 s

## Steps
1. Open Avatar Talk — Go to /app/avatar-talk and pick VEED Fabric 1.0 (the button on this page selects it) in Talking photo mode.
2. Add the image — Upload a clear, well-lit image of one face, up to 20 MB: a photo, an illustration or a character. Frame it the way you want the video to look.
3. Add the speech — Upload or choose a voice track (3 s to 5 min), or record one with your microphone. Clean speech without music gives the best lip sync.
4. Pick 480p or 720p and generate — 480p costs 21 credits a second, 720p costs 42. The exact price for your track shows before you generate, and the video lands in your library.

## In depth

### How Fabric works
Fabric builds the video from the still. VEED describes it as turning any image into a talking video with its own lip-sync technology, from a typed, recorded or uploaded voice [1][2]. It isn't a filter over existing footage. VEED put it in its editor on 15 September 2025 and opened the API on 25 September [3]. On API platforms it takes an image and an audio file and returns an MP4 at 480p or 720p [4].

### Fabric price next to other talking-photo engines
Per second of speech on Kenerate, in credits. All run in Avatar Talk.
- VEED Fabric 1.0: 21 at 480p, 42 at 720p, up to 5 min, min 3 s. 30 s at 480p = 630.
- InfiniteTalk: 9 at 480p, 18 at 720p, up to 10 min, one or two speakers.
- OmniHuman 1.5: 48, up to 60 s, full-body gestures.
- Sync Lipsync 3: 40.5, up to 5 min, the most exact mouth shapes.
- Re-voicing an existing video instead: VEED Lipsync 2 at 22.5, LatentSync at 3.

### Getting a clean Fabric result
Start from an image with one clearly lit face looking roughly at the camera, mouth visible and unobstructed. Use dry speech with no background music: the model follows the voice, and music muddies the mouth shapes. Draft at 480p (half the price), then re-run the keeper at 720p. For a long script, one 5-minute run keeps the same face and framing throughout, with no stitching.

References:
[1] VEED Fabric 1.0: VEED: https://www.veed.io/ai/fabric-1-0
[2] Introducing VEED Fabric: VEED blog: https://www.veed.io/blog/introducing-veed-fabric
[3] VEED adds Fabric 1.0: CGWorld (September 2025): https://cgworld.jp/flashnews/01-202509-VEED-Fabric.html
[4] veed/fabric-1.0 API: fal: https://fal.ai/models/veed/fabric-1.0

## FAQ
Q: What is VEED Fabric 1.0?
A: VEED Fabric 1.0 is VEED's AI talking-video model. You give it one still image and an audio track, and it generates a video of that subject speaking, with lips, head and expression following the speech. VEED added it to its editor in September 2025 and released an API the same month. It appears on model platforms as veed/fabric-1.0.

Q: How much does VEED Fabric cost on Kenerate?
A: 21 credits per second of speech at 480p and 42 at 720p, with anything under 3 seconds billed as 3. A 30-second talking video costs 630 credits at 480p or 1,260 at 720p. Credits come in one-time packs that never expire, from $15 for 1,300 credits. There's no subscription.

Q: How long can a Fabric video be?
A: Up to 5 minutes of speech per video on Kenerate. The video runs as long as your audio. VEED's own editor limits Fabric to 10 seconds a generation on the free tries and 30 seconds on paid plans, so longer pieces there are stitched from several clips.

Q: What resolution does VEED Fabric 1.0 make?
A: 480p or 720p. 480p is the default and costs half as much. If you need 1080p, Kenerate Avatar in the same studio renders 720p or 1080p presenter videos.

Q: Can Fabric animate a cartoon or an illustration?
A: Yes. VEED shows Fabric working on illustrations, anime characters and animals as well as photos. A clear, front-facing face with a visible mouth gives the most convincing lip sync.

Q: Is VEED Fabric free?
A: VEED gives every account 3 free Fabric generations of up to 10 seconds in its own editor. On Kenerate, Avatar Talk doesn't need a paid plan, but the smallest Fabric run (3 seconds at 480p) costs 63 credits, more than the 25 starter credits, so you'll need a credit pack.

Q: VEED Fabric vs VEED Lipsync: what's the difference?
A: Fabric makes a talking video from a still image. VEED Lipsync 2 takes a video you already have and re-syncs the mouth to new speech. On Kenerate, Fabric is in Avatar Talk's Talking photo mode (21 or 42 credits a second) and VEED Lipsync 2 in Lip sync a video mode (22.5 credits a second).

Q: VEED Fabric vs InfiniteTalk and OmniHuman: which should I use?
A: Fabric sits in the middle. InfiniteTalk is cheaper (9 credits a second at 480p, up to 10 minutes, one or two speakers). OmniHuman 1.5 costs 48 a second and adds full-body gestures for clips up to a minute. Fabric's strength is polished UGC- and ad-style talking heads from any image, up to 5 minutes.

Q: Does VEED Fabric need a text prompt?
A: No. Fabric takes only the image and the audio, and the speech drives the mouth, head and expression. Make the voice first, in Kenerate's voice studio or with your own recording, then pair it with the image in Avatar Talk.

Q: Can I use a photo of a celebrity or someone else?
A: Only with their permission. Use your own photo, a consenting person's photo or a fictional character. Making real people say things they didn't say without consent breaks Kenerate's rules and, in many places, the law.

## Related pages
- AI talking avatar: https://kenerateai.com/ai-talking-avatar
- Talking photo AI: https://kenerateai.com/talking-photo-ai
- Lip sync AI: https://kenerateai.com/lip-sync-ai
- InfiniteTalk: https://kenerateai.com/models/meigen-ai/infinitetalk
- OmniHuman 1.5: https://kenerateai.com/models/bytedance/omnihuman-1-5

## Sources
- VEED Fabric 1.0: VEED: https://www.veed.io/ai/fabric-1-0
- Introducing VEED Fabric: VEED blog: https://www.veed.io/blog/introducing-veed-fabric
- VEED adds Fabric 1.0: CGWorld (September 2025): https://cgworld.jp/flashnews/01-202509-VEED-Fabric.html
- veed/fabric-1.0 API: fal: https://fal.ai/models/veed/fabric-1.0