Text-to-image, editing and batch generation in one place. Write prompts in your own language — 14 languages understood natively. Built for creators and designers.
Get 100 free credits on signup — no credit card, no API key.
Image models HOTThe workhorse image model: native 2K output (2048×2048 at 1:1) and prompts understood directly in 14 languages.
◆ 20View model →Qwen-Image ProImage models PROFlagship: stronger instruction following and subject consistency
◆ 30View model →
Image modelsConversational editing: add an outfit, change the background, remove a bystander — one line each.
View model →
Video models NEWCurrent flagship video model with steadier camera work and timing, great for short films and ad clips.
◆ 360 / 5sView model →Wan 2.7 T2VVideo models NEWWan 2.7 flagship — native audio, multi-shot storytelling, 720P/1080P
◆ 150 / 5sView model →Wan 2.2 LiteVideo models LITEWan 2.2 lite — silent video (no audio track), 480P, cheapest text-to-video
◆ 30 / 5sView model →
Sample promptNeon-lit cyberpunk city at night, rain-soaked streets, a woman in a translucent raincoat, cinematic teal and magenta lighting, ultra detailed, 2K
Sample promptTraditional Chinese ink wash painting of misty mountains, golden foil accents, a winding river, minimalist composition, rice paper texture, 2K
Sample promptStudio product photography of a glass perfume bottle on a marble podium, softbox lighting, subtle reflections, clean beige background, commercial shot
Sample promptCute 3D cartoon capybara mascot, chibi style, soft clay render, pastel colors, big round eyes, studio lighting, blind-box toy design



100 free credits on sign-up, no credit card required
New accounts get 100 free credits