VideoEnhancerVideoEnhancer
VideoEnhancerVideoEnhancer

AI-powered video enhancement platform. Upscale and compress your videos with state-of-the-art models.

Product
  • AI Video Generator
  • Image to Video
  • AI Video Upscaler
  • Image Upscaler
  • Image Compare
  • Short Video HD
  • Kling Video HD
  • Video Compress
  • Discord Compressor
  • Video Restoration
  • SparkPix
Video models
  • Seedance 2.5
  • Seedance 2.0
  • Wan 3.0
  • Wan 2.7
  • Wan 2.2
  • Grok Imagine Video 1.5
  • MiniMax H3
Company
  • About Us
  • Pricing
  • Contact
Legal
  • Terms of Service
  • Privacy Policy
Partners
Featured on Fazier
Friends
  • AiTop10 Tools Directory
  • AIAI Tools
  • MeoAI

© 2026 VideoEnhancer. All rights reserved.

TermsPrivacyContact

AI Video Editor

How it works
Model
Speed
Image · optional
Video · optional
A reference clip, billed at the output rate on top of the output; input + output at most 30s.
Prompt
0/2000
Duration
Aspect ratio
Resolution
Audio

Wan 3.0

Alibaba’s Wan 3.0: up to 30 seconds in a single shot with audio, at standard or prime speed, from 7 credits a second. 35 credits for five seconds at 480P.

Wan 3.0, image to video from our demo still. Standard 480P, 5 seconds, 35 credits, under four minutes. Prompt: “Gentle camera push-in; hair and fabric move in a light breeze; soft natural light.”

Wan 3.0

Wan 3.0 is Alibaba’s video generation model that produces clips of up to 30 seconds with their own audio, from a prompt, from a first and last frame, or from reference images and video. Use it when the take has to run long in one continuous shot and you want the cheapest 30 seconds in the studio.

It runs here at standard speed for 7 credits a second at 480P, 14 at 720P and 29 at 1080P, so a 5-second 480P clip is 35 credits and a 30-second 1080P one is 870. A prime tier at roughly 1.5 times the rate returns the same clip several times faster. The Create button shows the exact figure before anything is charged.

Two speeds, one model

Standard is the everyday tier: 7, 14 or 29 credits a second by size. Prime is the same model with the queue skipped: 11, 22 or 45 credits a second, and the vendor quotes it at about six times the speed. Our standard-speed jobs came back in two and a half to four minutes; the same 5-second 480P clip on prime came back in 80 seconds for 55 credits instead of 35. Prime earns its premium when you are waiting on the result to make the next decision, not for a take you will collect later.

Wan 3.0 takes 4:3 and 3:4 frames as well as 16:9, 9:16 and 1:1, and it accepts a reference video alongside a prompt. That footage is billed at the same per-second rate as the output, on top of it: 10 seconds generated from 8 seconds of reference is charged as 18.

What it costs

Per second of billed video. Reference video input counts toward the seconds; images, audio and documents do not.

SizeStandardPrime5 s standard30 s standard
480P7 cr/s11 cr/s35210
720P14 cr/s22 cr/s70420
1080P29 cr/s45 cr/s145870

Our test

The hero clip is an image-to-video job from the still on our home page: standard speed, 480P, 5 seconds, one sentence of prompt. It cost 35 credits, took under four minutes, and came back as an 854×480 MP4 of 5.0 seconds with audio. A text-to-video job at the same settings, run the day before, took about two and a half minutes.

The same still and sentence went through Wan 2.2, Wan 2.7 and Seedance 2.5, so the four image-to-video results can be compared on one shot.

A third job on 2026-09-24 attached a 5-second reference clip alongside a prompt at standard 480P for 5 seconds: billed as 10 seconds at 7 credits, 70 credits, back as an 854×480 clip with audio.

FieldValue
InputOne still, 1280×720, plus the prompt in the caption
SettingsStandard, 480P, 5 seconds, 16:9
Cost35 credits (from $0.35)
Wall timeAbout 3½ minutes; 2½ for the text-only run
Output854×480, 5.0 s, MP4 with audio
Reference clip 5 s + 5 s output70 credits (billed as 10 s at 7)

Wan 3.0, Wan 2.7 or Wan 2.2?

Here 3.0 is the only one of the three that starts from text, reaches 30 seconds, or offers 1080P at standard speed for less than 30 credits a second. 2.7 runs here as image to video only, 2 to 15 seconds, and its distinguishing feature is an audio input that drives the motion. 2.2 is the cheapest way to see whether a still moves the way you expect: 60 credits, silent, back in about a minute.

How to run Wan 3.0 here

  1. 1. Start from a still, or from the prompt alone Add a picture to animate it (an ending frame is optional), or leave it out and the clip is generated from the prompt.
  2. 2. Describe the motion One sentence: what moves, how the camera moves, the light.
  3. 3. Set size and length Start at standard 480P and 5 seconds — 35 credits. Go long only once the motion is right.
  4. 4. Create Two to four minutes at standard speed; the clip waits in History.

When not to use this

  • Quick prompt drafts. 35 credits and three minutes a draft is still slower than Grok Video 1.5 at 25 credits in about a minute; find the prompt there.
  • Motion driven by a music track. That is Wan 2.7’s audio-input feature; Wan 3.0 generates its own sound but does not follow yours.
  • Free reference footage. Any video you attach is billed second for second alongside the output.
  • Readable text in frame, which comes out as plausible nonsense on every model here.
  • Footage of a real person who has not agreed to it.

Questions people ask

How much does Wan 3.0 cost, and is it free?

At standard speed, 7 credits a second at 480P, 14 at 720P and 29 at 1080P; prime speed is 11, 22 and 45. One credit is one US cent, so a 5-second 480P clip is 35 credits and a 30-second 1080P one is 870. It is not free: 35 credits is more than the 25 a new account gets. The figure is on the Create button before the job starts.

Is Wan 3.0 open source, and can it run locally?

No. Alibaba offers Wan 3.0 as an API on its Model Studio, has published no weights for it, and has announced no date for doing so (checked 17 November 2026). The newest Wan with downloadable weights is Wan 2.2, under Apache 2.0. This page calls the Wan 3.0 API for you.

What is Wan 3.0 prime?

The same model with priority processing, billed at roughly one and a half times the standard rate. At standard our jobs took two and a half to four minutes; the same clip on prime took 80 seconds, 55 credits against 35 at 480P. It pays when you are waiting on the result to decide the next step.

How do I use a reference with Wan 3.0?

Put the clip in the Video slot and, if a person or object from a photo should appear in it, the photo in the Image slot; with a clip attached the photo is read as a reference, not as the opening frame. Call them “video 1” and “figure 1” in the prompt. The clip’s seconds are billed with the output at the same rate and the two may not pass 30 seconds together: our 5-second clip plus a 5-second result at 480P came to 70 credits.

Can Wan 3.0 swap a character in a video?

Yes. Attach the clip and a photo of the new person, and write who is replaced, who replaces them and what must not change. The model keeps the movement and the camera and redraws the person from the photo. An 8-second clip turned into an 8-second result at 720P is 224 credits; the replace character page has the prompts.

Is Wan 3.0 better than Seedance 2.5?

We have not run a blind comparison, so we do not rank them. They overlap: both reach 30 seconds in one take with audio, and both can pick their own length. Wan 3.0 is cheaper (7 against 19 credits a second at 480p), offers 1080P here and has the faster prime tier. Seedance 2.5 takes more reference material, up to 30 images, 10 videos and 10 audio clips, and can extend a clip you already have.

Related tools

AI Car Driving VideoThe viral trend on Wan 3.0: your photo in a luxury driving clip.Replace CharacterWan 3.0 with a photo and a reference clip: swap who appears, keep the shot.Wan 2.7Image to video with an audio track that drives the motion.Wan 2.2The 60-credit, one-minute check that a still moves right.Seedance 2.5The other 30-second model.Text to videoEvery model that generates from a prompt alone.

Checked 2026-11-17.