Vidu Image Q2 is a reference-to-image AI model for designers and creators who need a character, product, or IP to stay identical across many generated pictures. Offered through Tencent's TokenHub platform (model id vidu-image-q2) and Vidu's own API (viduq2), it extends the earlier Q-series with sharper consistency and 4K output.
Core Features
- Multi-reference input: accepts 0 to 7 reference images and preserves the subject's face, clothing, pose, and texture across outputs.
- Flexible ratios and resolution: 16:9, 9:16, 1:1, 3:4, 4:3, 21:9, 2:3, 3:2, or auto, at 1080p, 2K, or 4K.
- Precise text and UI rendering: renders Chinese and English text cleanly and reproduces UI and chart details at pixel level.
- Spatial understanding: reconstructs environment geometry so subjects move naturally through scenes instead of clipping.
- Fast generation: reference-to-image results in as little as about 5 seconds, with image editing and text-to-image in the same endpoint.
Use Cases / Best For
- Brand and IP studios placing one mascot into posters, packaging, and ads.
- Comic and short-drama teams keeping a character consistent across shots.
- E-commerce and marketing teams producing on-brand product visuals at scale.
Pricing
Vidu Q2 is billed through Tencent TokenHub and the Vidu API on a per-task basis. The platform has offered limited free trials, with paid usage for higher volume and 4K renders; exact per-image pricing varies by resolution and plan.
Our Take
Best for anyone whose work lives or dies on subject consistency; the trade-off is that pure artistic stylization still trails dedicated art models like Midjourney. See more in our Image Generation tools list.
