Wan Video 2.1 is Alibaba's open-source video model, released in February 2025 under Apache 2.0. It generates video from a text prompt or a still image, and it was the release that made high-quality open-weight video generation practical on ordinary hardware.
In Phygital+ it runs as one of the version options in the Wan Video node, with turbo and plus variants, resolutions from 480P to 1080P, start and end frame guidance, and a negative prompt for the artefacts you want suppressed.
Social drafts, concept clips, promo visuals, and animated product or portrait stills
Social, content, and prototyping teams that need volume and speed over flagship fidelity
Open-weight video generation with start and end frame control, cheap enough to iterate on.
Wan 2.1 shipped under Apache 2.0, which permits commercial use, fine-tuning, and redistribution without additional terms. That is unusual among video models of this class, and it is why a large community of LoRAs and workflows grew around it.
Connect a still to Start Frame and the clip opens on a composition you already approved. Add an End Frame when the final shot matters and the model works out the progression between them.
Five, ten, or fifteen seconds at 480P, 720P, or 1080P. Test motion at the small end where a run is cheap, then re-render the take that works at delivery size instead of gambling on a full-quality first attempt.
Blur, flicker, warped anatomy, unwanted text: name what should not appear and the model suppresses it. Concrete exclusions work better than vague quality complaints.
The Wan Video node also carries 2.2, 2.5, 2.6, and 2.7. Same canvas, same controls, so moving to a newer version when a project needs reference images or clip continuation is a dropdown change, not a rebuild.
Wan Video 2.1 fits wherever you need a lot of short clips quickly: social drafts, concept tests, and animated stills that do not justify a flagship render.
Pick the version, write a motion-aware prompt, optionally anchor a start frame, and generate.
Add a Wan Video node and select 2.1 turbo for faster iteration or 2.1 plus for higher quality. Write a prompt covering subject, setting, action, camera movement, lighting, and pacing.
Connect a Start Frame to lock the opening composition, and an End Frame if the final shot matters. Set duration, resolution, and aspect ratio, and use the negative prompt for concrete exclusions such as blur, flicker, or warped anatomy.
Fix a seed to reproduce a take. Test at short duration and lower resolution first, then re-run the version that works at delivery size. Chain into a video upscaler or an audio node, then export with commercial rights on your plan.
One subscription across 30+ AI models, with no per-tool credit balances or separate signups. Credit cost per generation is shown live in the node before you run it.
500 weekly credits to test the model and see the output quality. No credit card required.
45,000 monthly credits with cheaper per-credit pricing, commercial rights, and video download.
90,000 monthly credits, up to 10 seats, shared workspace, centralized billing, and priority support.
210,000 monthly credits, API and integration support, unlimited seats, and dedicated management.
Full specs for Wan Video 2.1 in Phygital+.
How Wan Video 2.1 compares with the other AI video models in the Phygital+ catalog.
Chain image generation, video, upscaling, and audio into one repeatable workflow.
Generate a first frame with an image model, then animate it in Wan 2.1 so the composition is approved before any motion is rendered.
Take a 480p or 720p Wan clip into a video upscaler to reach delivery resolution without paying for a higher-res generation.
Generate narration or sound design in an audio node and lay it under the clip, or connect audio as a rhythm reference at generation time.
Write the scene description and shot list with a text node, then send it into the prompt field.
Everything you need to know about Wan Video 2.1 in Phygital+.
Wan 2.1 is Alibaba's open-source video generation model, released in February 2025 under the Apache 2.0 licence. It was the first version of the Wan series and covers text-to-video and image-to-video, built on a diffusion transformer architecture with a purpose-built spatio-temporal VAE.
It shipped in two sizes: a 14B model for quality and a 1.3B model light enough to run on consumer hardware. Alibaba open-sourced four Wan 2.1 models in February 2025, covering text-to-video plus image-to-video at 480p and 720p, and later added a first-and-last-frame variant.
In Phygital+ the Wan Video node exposes 2.1 turbo and 2.1 plus alongside newer versions up to 2.7. Turbo is the faster option for iteration; plus trades speed for quality. Newer versions add reference images, video references, and clip continuation, which 2.1 does not support.
The node offers 480P, 720P, and 1080P, with durations of 5, 10, or 15 seconds and aspect ratios of 1:1, 16:9, 9:16, 4:3, and 3:4. Shorter and smaller is the right starting point: test the motion first, then re-run at delivery size once the clip works.
Yes. Connect a Start Frame to set the opening composition and an End Frame to steer where the clip lands. Both are optional, and the start frame is the usual way to animate an approved still rather than gambling on a text prompt.
Enhance Prompt expands a short brief into a fuller cinematic description before generation. Leave it on when you want the model to fill in atmosphere and detail. Turn it off with a short, precise prompt when you need strict adherence to exactly what you wrote.
Describe the subject, the setting, the action, the camera movement, the lighting, and the pacing, in roughly that order. Say explicitly what moves and what stays still when animating a photo. Use the negative prompt for concrete exclusions such as blur, flicker, or warped anatomy rather than vague quality complaints.
Usage is credit-based and included in every plan, from Free up to Enterprise. The credit cost per generation is shown in the node before you run it, and no GPU or local setup is needed.
Outputs created on any paid plan come with commercial-use rights. Wan 2.1 itself is published under Apache 2.0, which permits commercial use, fine-tuning, and redistribution at the model level.
Browse the full catalog of 30+ AI models available in Phygital+.
Join 100+ teams using Phygital+ – every model in one workspace
Try Phygital+ free
