Limited-time offer: save up to 30% on subscriptions
Upgrade NowThe thinking and improvements behind every release
One idea, more ways to bring it to life. GPT Image 2.6 adds looping GIF animations and SVG vector output, along with transparent-background assets and image creation at up to 4K. From product posters and social GIFs to brand icons, create in more formats.
Bring ideas to life: Generate looping GIFs from text prompts for reactions, social posts, and brand animations.
Scale without losing sharpness: Create SVG icons and illustrations that stay sharp at any scale for websites, apps, and branding.
Make assets easier to reuse: Generate transparent PNGs for product showcases, stickers, and creative layouts.
Create in sharp detail: Output images at up to 4K or edit uploaded reference images for product shots, ad posters, and other visuals.
Stunning at first glance. Even better up close. Nano Banana 2.1 brings a new level of image quality, with natural lighting and fine textures for greater realism and visual expression. From portrait photography and product ads to creative posters, enjoy clearer, more vivid and striking visuals.
Better image quality, more striking detail: finer skin, hair, fabric and metal textures make every detail clearer and richer.
Realistic lighting, more vivid images: more natural transitions between light and shadow, with richer tonal depth, bring greater realism and a cinematic feel to your visuals.
4K resolution reveals more detail: supports 1K, 2K and 4K output with improved rendering in ultrawide formats for better commercial photography, brand posters and advertising visuals.
Turn an idea straight into a high-quality image. GPT Image 2.5 delivers more realistic lighting and textures, more precise image editing, and faster generation, so every creation is more consistent and controllable.
Up to 50% faster generation: create and batch-generate images faster with far less waiting.
High-quality 4K output: finer textures, more natural lighting, and more lifelike visuals.
Stronger prompt understanding: handles complex style, layout, poster, product shot, and UI design requests more accurately.
Turn an idea straight into a complete video. Wan 3.0 can generate videos up to 30 seconds long while keeping characters, motion, camera work, and sound consistent throughout longer videos.
Up to 30 seconds per generation: create fuller shots, scenes, and stories in a single pass.
More realistic physics: cloth, fluids, collisions, and multi-object motion rendered naturally.
Stronger visual continuity: characters, objects, and scene details stay consistent throughout longer videos.
Smooth 30 fps output: get more natural, fluid motion.