{"generate":{"model":{"gpt-image-2":{"description":"Realistic images with readable multilingual text and precise edits. Up to 4K, 16 reference images."},"gpt-image-2-5-flare":{"description":"Everyday default, faster than GPT Image 2. Holds one character and style across grid sheets; up to 4K."},"background-remover":{"description":"Cuts out the subject in one click. Export on transparent, white, green, or blurred backgrounds."},"image-upscaler":{"description":"Enlarges photos 2×, 4×, or 8× for print or crisp display."},"image-extender":{"description":"Changes a photo's aspect ratio without cropping; AI fills in the extra scene. From 21:9 to 9:21."},"video-upscaler":{"description":"Upscales any video to 1080p, 2K, or 4K with sharper textures and cleaner edges."},"video-background-remover":{"description":"Removes a video's background and keeps the original audio. Export transparent or swap in a solid color."},"lip-sync-video":{"description":"Makes a portrait photo talk, lip-synced to your audio file. 720p or 1080p; length follows your audio."},"motion-control":{"description":"Copies the motion from a reference video onto your character image. Up to 30s, 720p or 1080p."},"grok-imagine":{"description":"Fast, cinematic images for brainstorming and trying styles. Edits with up to 5 reference images."},"grok-imagine-video-1-5":{"description":"Fast video drafts from text or a first frame. 480p or 720p, clips of 1–15s."},"nano-banana":{"description":"Quick everyday images and edits for casual experiments. Blends up to 10 reference images."},"nano-banana-2":{"description":"Balanced speed and quality, the go-to Google pick. Up to 4K; edits with up to 14 reference images."},"nano-banana-2-1":{"description":"Newest Nano Banana from Google. Up to 4K, extra-wide 8:1 ratios, edits with up to 14 references."},"nano-banana-2-lite":{"description":"Fastest Nano Banana tier for 1K drafts, with legible text and consistent characters."},"nano-banana-pro":{"description":"Studio-grade accuracy for complex scenes, using real-world knowledge. Up to 4K, 8 reference images."},"turbo-image":{"description":"Very fast drafts with good prompt following and text rendering. Edits with up to 10 reference images."},"seedance-2":{"description":"Top Seedance tier with sound, the only one up to 4K. Text, start and end frames, or image, video, and audio references; 4–15s."},"seedream-5-lite":{"description":"Fast, lower-cost photoreal images and edits at 2K or 3K. Up to 14 reference images."},"seedream-5-flash":{"description":"Fast image creation and editing, up to 2K."},"seedance-2-fast":{"description":"Faster, lower-cost Seedance 2 with sound and the same frame and reference inputs. 480p or 720p, 4–15s."},"veo3":{"description":"Highest-quality Veo 3.1 tier for finals. 8s clips with native sound, 720p to 4K, optional start and end frames."},"veo3-fast":{"description":"Faster, lower-cost Veo 3.1 for iterating. 8s clips with sound, up to 4K; guide with frames or 3 reference images."},"veo3-lite":{"description":"Budget Veo 3.1 tier for quick drafts. 8s clips with sound, up to 4K, with optional start and end frames."},"image-watermark-remover":{"description":"Erases text and logo watermarks from a photo."},"video-watermark-remover":{"description":"Erases watermarks, logos and captions from clips up to 60 s."}}}}