Génération de vidéos

Wan 3.0

par Alibaba

Alibaba's latest video model, in public beta.
Native 30s clips from prompts, docs or webpages.

Wan 3.0

Fonctionnalités clés

Un travelling avant ininterrompu dans une ruelle luisante de pluie, les néons défilantCommencer à créer

Features confirmed from Alibaba's Wan 3.0 public-beta launch, August 2026.

Spécifications techniques

Up to 30s

Native 30 seconds in a single run, with intelligent duration control.

480p to 1080p

Three tiers: 480p, 720p, and 1080p. There is no 4K output.

Public beta

Live on Alibaba Cloud Model Studio and Qwen Cloud as wan3.0-video.

API only

No open weights. Alibaba's open Wan line stops at Wan 2.2.

Cas d'utilisation

Report to video

Hand Wan 3.0 a deck, spreadsheet, or pdf and it builds a video sequence from the data and text inside it.

Full ad spot in one take

A 30-second clip covers a broadcast spot in a single generation, with no cuts to stitch or match afterwards.

Reference-locked series

Omni-Reference holds a character, product, or set across every shot, so a sequence reads as one piece.

Exemples de prompts

Long single take

One unbroken push down a rain-slick alley, neon reflections sliding past

Edit prompt

Reference continuity

The same courier crosses three city blocks, morning haze to midday glare

Edit prompt

Product reveal

Studio-lit rotation of a brushed steel kettle on a dark stone slab

Edit prompt

Character performance

A violinist finishes a phrase, lowers the bow, and exhales in low light

Edit prompt

Landscape reveal

Slow aerial pull-back from a river delta as the fog burns off at sunrise

Edit prompt

In-frame typography

Coastal cliff at golden hour with a clean lower-third title card

Edit prompt

Tarifs simples

Commencez gratuitement dès aujourd'hui, avec la possibilité de mettre à niveau ou d'annuler à tout moment.

Basic

$9/ mois
facturé comme $0 par an

900 mensuel crédits

1 utilisateur uniquement

Tous les modèles

Workflows

Standard

$24/ mois
facturé comme $0 par an

3200 mensuel crédits

1 utilisateur uniquement

Tous les modèles

Workflows

Pro

$45/ mois
facturé comme $0 par an

6200 partagé mensuel crédits

1 utilisateur

+ jusqu'à 4 plus à un coût supplémentaire

Tous les modèles

Workflows

Pro Max

$170/ mois
facturé comme $0 par an

24000 partagé mensuel crédits

1 utilisateur

+ jusqu'à 9 plus à un coût supplémentaire

Tous les modèles

Workflows

Enterprise

Pour des limites plus élevées

Personnalisé

conditions de tarification et de facturation

Crédits à haut volume
Limites de sièges personnalisées
Tous les modèles
Workflows
Pricing Gradient

Free

Pour expérimenter

$0

gratuit pour toujours

Jusqu'à 20 crédits
1 utilisateur uniquement
Modèles limités
Workflows
Comparer les détails des plans

Questions fréquentes

What is Wan 3.0?
Wan 3.0 is Alibaba Tongyi Lab's latest video model, in public beta since 6 August 2026. It generates native 30-second clips at up to 1080p with audio in the same pass. Its signature feature is Omni-Reference: alongside text, image, audio, and video, it accepts documents, spreadsheets, slides, pdfs, and webpages and builds video from them. It is available through Alibaba Cloud Model Studio and Qwen Cloud as the model wan3.0-video.
Has Wan 3.0 been released?
Yes. Alibaba launched Wan 3.0 in public beta on 6 August 2026, on Alibaba Cloud Model Studio and Qwen Cloud, with wider access on wan.video rolling out to members. The model ID (wan3.0-video), the resolution tiers, and per-second API pricing are all now published, so it is a live model rather than the rumor it was through July 2026.
How long are Wan 3.0 clips?
Up to 30 seconds in a single run, with intelligent duration that matches clip length to the prompt so short scenes are not padded. That is double the 15-second ceiling on Wan 2.7, and it is built for one continuous take rather than short fragments stitched together.
Is Wan 3.0 open source?
No. Wan 3.0 is a closed, API-only model, and no weights have been published on Hugging Face, GitHub, or ModelScope. Alibaba's last open-weight video flagship remains Wan 2.2 under Apache 2.0, released in 2025. If self-hosting is a requirement, Wan 2.2 is still the newest downloadable Wan, and any site advertising Wan 3.0 weights or a free download is not offering the real model.
What can Wan 3.0 create from a document?
Wan 3.0 accepts doc, xls, ppt, pdf, txt, key, pages, numbers, and md files, plus webpage URLs. It reads the content and generates a video sequence from it, which Alibaba describes as turning static, text-heavy data into finished video. This document- and webpage-to-video capability is the main thing that sets Wan 3.0 apart from other current video models.
How does Wan 3.0 compare with Seedance 2.5?
Both launched within days of each other in early August 2026, and both generate 30-second single-pass clips with audio. Seedance 2.5 accepts more reference inputs, up to 50 across images, video, and audio, while Wan 3.0's distinguishing capability is building video from documents and webpages. Neither model has independent benchmarks yet, and both are closed, API-only releases, so any head-to-head today rests on demos rather than measured scores.

Autres modèles

MiniMax H3 Max

fal.ai

MiniMax H3, post-entraîné par fal. Plus fidèle au brief, reconstruit pour la vitesse.

Grok Imagine Image 2.0

xAI

Le modèle d'image de précision de xAI. Typographie de designer, retouche ciblée, Smart Resize.

Seedance 2.5

ByteDance

Le modèle vidéo nouvelle génération de ByteDance. Jusqu’à 30 s en 1080p natif, 50 références, audio natif.

Kling 4.0

Kling

Prochain modèle phare Kling de Kuaishou, attendu bientôt. Clips plus longs, plus de détail.