Wan AI 3.0 Video: Text & Image to Video Online
Bottom line
Alibaba's newest Wan: one model for text or image input, native audio, clips from 2 to 30 seconds. Start from a prompt or a still image in the Text to Video tool; VdoBloom shows the exact credit cost before each Wan AI 3.0 Video run, with no separate Wan AI account.
Wan 3.0 Video is the newest generation in Alibaba's Wan line on VdoBloom, released August 2026, and it collapses what used to be separate text-to-video and image-to-video models into one: it handles both from the same settings panel. Every second from 2 to 30 is selectable, at 480p, 720p or 1080p, in five aspect ratios (16:9, 4:3, 1:1, 3:4, 9:16), and it generates audio natively alongside the picture.
That 30-second ceiling is double what Wan 2.6 or 2.7 offer (both cap at 15 seconds), making Wan 3.0 the option when a shot needs to actually develop rather than land in a single beat. Pricing scales per second and resolution: roughly 3.2-3.5 credits/second at 480p, 6.4-6.7 at 720p, and 12.8-13 at 1080p -- at a matching 5-second 720p clip, that's the same 32 credits as Wan 2.7 Text-to-Video, so the longer ceiling doesn't come at a per-second premium over the previous generation.
It carries VdoBloom's HOT content label, the platform's most permissive filtering tier, consistent with the rest of the Wan family. As with the other recent Wan releases, Alibaba has not published Wan 3.0's weights, so VdoBloom's hosted access is the way to run it -- no GPU or install required. The exact credit cost for your chosen duration and resolution is shown before you generate.
What it costs on VdoBloom
Wan 3.0 bills per second across three resolution tiers: roughly 3.2-3.5 credits/second at 480p, 6.4-6.7 at 720p, 12.8-13 at 1080p. Worked examples at 720p: 2s = 13 credits, 5s = 32 credits, 10s = 64 credits, 30s = 192 credits. At 1080p: 2s = 26 credits, 5s = 64 credits, 10s = 128 credits, 30s = 384 credits.
Wan 3.0 vs Wan 2.7: same price, longer ceiling
At a matching 5-second 720p clip, Wan 3.0 and Wan 2.7 Text-to-Video both cost 32 credits -- the newer model's extended range doesn't carry a per-second premium at that duration. Where they differ: Wan 3.0 runs up to 30 seconds (versus 2.7's 15), adds a 480p tier, generates audio for both text and image starts, and handles image-to-video in the same model. Wan 2.7 Image-to-Video keeps a feature Wan 3.0 doesn't list -- first- and last-frame control for planned transitions. Pick Wan 3.0 for longer or audio-native work across either input type; pick Wan 2.7 Image-to-Video specifically when you need to pin an exact ending frame.
What creators use Wan AI 3.0 Video for
- Generate a full 30-second scene with sound in one pass, well past what most text-to-video models on the platform allow.
- Switch between text-only and image-start generation in the same model without picking a different tool.
- Use the 480p tier for quick, cheap drafts of a long-form idea before committing to 720p or 1080p.
- Produce native-audio clips for social platforms without a separate sound-design step.
How to use Wan AI 3.0 Video on VdoBloom
- 1Open the Text to Video tool and select Wan AI 3.0 Video from the model picker.
- 2Write your prompt — or upload a starting image for image-to-video.
- 3Choose your duration (2–30s), quality (480p/720p/1080p), aspect ratio, review the credit cost, and generate.
Frequently asked questions
How much does Wan 3.0 cost on VdoBloom?
It scales per second and resolution: roughly 3.2-3.5 credits/second at 480p, 6.4-6.7 at 720p, 12.8-13 at 1080p. Worked examples: a 5-second 720p clip is 32 credits, a 10-second 720p clip is 64 credits, a 30-second 720p clip is 192 credits; at 1080p those are 64, 128 and 384 credits.
Do my 10 free signup credits cover a Wan 3.0 render?
For the shortest, cheapest option, yes: a 2-second 480p clip costs 7 credits, leaving 3 of your 10 free credits. Anything longer or at a higher resolution will need more credits than the free starting balance.
Is Wan 3.0 open source?
As of this writing, Alibaba has not released Wan 3.0's weights the way it did for the earlier Wan 2.1/2.2 generations. It's accessed through hosted platforms like VdoBloom.
Can I run Wan 3.0 locally?
No -- without published weights there's no local install path. VdoBloom's hosted version is how you access Wan 3.0 without a GPU or setup.
What's the longest clip Wan 3.0 can generate?
30 seconds, in one-second increments starting at 2 seconds. That's double the 15-second cap on Wan 2.6 and 2.7, making 3.0 the pick for anything that needs to develop over more than a single beat.
Does Wan 3.0 generate audio?
Yes, natively, alongside the video -- consistent with Wan 2.6 Flash and Wan 2.7 Text-to-Video, but now available across the full 2-30 second range and both text and image input.
How does Wan 3.0 compare to Wan 3.0 Prime?
They share the same durations, resolutions and aspect ratios. Prime runs faster generation at a higher price -- roughly 4.9-5 credits/second at 480p, 10-10.5 at 720p, and 20-20.5 at 1080p, against the base model's 3.2-3.5, 6.4-6.7 and 12.8-13. A 5-second 720p clip is 32 credits on the base model versus 51 on Prime.
How strict is Wan 3.0's content filter?
VdoBloom rates it HOT, the platform's most permissive tier, matching the rest of the Wan lineup. Platform-wide rules on illegal content still apply, and a rejected prompt isn't billed.
Does Wan 3.0 replace Wan 2.7?
It's not a strict replacement -- both remain available. Wan 3.0 adds a longer duration ceiling, a 480p budget tier, and one model for both text and image input; Wan 2.7 Image-to-Video keeps its first/last-frame control, which Wan 3.0's spec doesn't document. For a clip under 15 seconds where endpoint framing matters, 2.7 Image-to-Video is still the more specific tool.