AI Briefing
KO

AI Gateway Adds Support for Alibaba's Wan Model

·2026.02.19 22:00

Key point

Alibaba's **Wan model** has been added to **AI Gateway**, supporting high-quality video generation using text and images.

Details

Alibaba's Wan model has been released through AI Gateway. Users can programmatically generate videos using the generateVideo function in AI SDK 6, or experiment with the model and download results in the Gateway Playground without coding.

The Wan model delivers smooth motion and high consistency, and supports the following key features:

  • Character Reference (R2V): Extracts a character's appearance from a reference video or image to generate new scenes
  • Flash Variants: High-speed generation for fast iteration
  • Flexible Resolutions: Supports 480p, 720p, and 1080p resolutions

The video generation feature is currently in beta and is available to Pro and Enterprise plan users as well as paid AI Gateway users. Supported models include Text-to-Video (t2v), Image-to-Video (i2v), and Reference-to-Video (r2v).

vercelvideo-generationalibabaai-gatewayai-sdkwan-models
Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Related

Vercel BlogTako Search is free on AI Gateway until September 30Vercel BlogVercel AI Gateway adds Pixel Canary model in stealth mode with free accessOpenAI BlogHiggsfield AI Launches New Video Features in One Day Using GPT-6 AstraElevenLabsElevenLabs Launches Image, Video, and Templates APIs for Automated Creative ProductionHacker NewsOpenComputer Hosts 'Ship Video' Agent Using Claude Opus 5.5 to Generate Explainer Videos from CodeTLDR AIPerplexity: AI Agents Bypass SPACE Network Policies; Direct Tests Reveal Weaknesses in 7 of 9 Third-Party Sandboxes