AI MODEL

WAN 2.5 T2I

Wan 2.5 Preview text to image is a mode from Alibaba's Wan 2.5 family, generating up to four images per request directly from language prompts. It supports variable aspect ratios between 1:4 and 4:1, with total pixel counts ranging roughly between 768 by 768 and 1440 by 1440. Outputs span a range of visual styles from photorealistic scenes to illustrative or stylized renders. The model emphasizes fidelity to prompt semantics, making sure scene elements, composition, subject relationships, and lighting line up with the user's description. Wan 2.5 T2I shares its multimodal backbone with the Wan video family, which contributes to strong structural coherence and reliable prompt adherence across repeated generations for marketing, editorial, and product imagery.

Release Date

September 24, 2025

Developer

Model Type

WAN 2.5 T2I Prompts & Outputs
Latest Models Added
  • Nano Banana 2 T2I
    February 26, 2026
  • FLUX 2 Klein 9B T2I
    January 15, 2026
  • GLM Image T2I
    January 13, 2026
  • LTX-2-19b
    January 6, 2026