# Nano Banana 2.1 configured models and workflows Image workspace: https://nanobanana21ai.app/ai-image-generator; homepage default: Nano Banana 2.1 at https://nanobanana21ai.app/. Video workspace: https://nanobanana21ai.app/ai-video-generator. Prices: https://nanobanana21ai.app/pricing.md. This catalogue comes from studio configuration, not marketing-page counts. Models shown in the selector have connected generation code paths; provider credentials, provider availability and valid mode-specific inputs are still required. It is not an uptime or output-quality guarantee. No private provider endpoint or credential is published. ## Image generation and editing ### Seedream 5.0 Pro - Public model ID: seedream-5-pro; provider brand: Seedream. - Modes: layer-decomposition, text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 5000 characters. - Resolutions: 1K, 2K; default: 2K. - Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation; maximum 31457280 bytes per image. - Output count: 1–6; format: selected workspace/model default. - Layer resolutions: auto, 1K, 1.5K, 2K; types: image/jpeg, image/png, image/webp, image/bmp, image/tiff, image/gif. ### GPT Image 2.5 Flare - Public model ID: gpt-image-2.5-flare; provider brand: OpenAI. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 32000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: 1:1, auto, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, 2:1, 1:2, 3:1, 1:3, 21:9, 9:21, 5:4, 4:5; editing geometry can follow the input image. - Reference image ceiling: 16; types: see workspace upload validation. - Output count: 1–10; format: selected workspace/model default. - Additional configured options: {"quality":{"values":["auto","low","medium","high","xhigh","max"],"default":"high"},"background":{"values":["auto","transparent","opaque"],"default":"auto"},"outputFormat":{"values":["png","jpeg","webp"],"default":"png"},"imageSize":{"presets":["auto","square_hd","square","portrait_4_3","portrait_16_9","landscape_4_3","landscape_16_9"],"multiple":16,"maxEdge":3840,"minPixels":655360,"maxPixels":8294400,"maxAspectRatio":3},"maskOptionalInEdit":true}. ### GPT Image 2.5 Sunburst - Public model ID: gpt-image-2.5-sunburst; provider brand: OpenAI. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 32000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: 1:1, auto, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, 2:1, 1:2, 3:1, 1:3, 21:9, 9:21, 5:4, 4:5; editing geometry can follow the input image. - Reference image ceiling: 16; types: see workspace upload validation. - Output count: 1–10; format: selected workspace/model default. - Additional configured options: {"quality":{"values":["auto","low","medium","high","xhigh","max"],"default":"high"},"background":{"values":["auto","transparent","opaque"],"default":"auto"},"outputFormat":{"values":["png","jpeg","webp"],"default":"png"},"imageSize":{"presets":["auto","square_hd","square","portrait_4_3","portrait_16_9","landscape_4_3","landscape_16_9"],"multiple":16,"maxEdge":3840,"minPixels":655360,"maxPixels":8294400,"maxAspectRatio":3},"maskOptionalInEdit":true}. ### GPT Image 2 - Public model ID: gpt-image-2; provider brand: OpenAI. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 20000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: auto, 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, 2:1, 1:2, 3:1, 1:3, 21:9, 9:21, 5:4, 4:5; editing geometry can follow the input image. - Reference image ceiling: 16; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Nano Banana 2.1 - Public model ID: nano-banana-2.1; provider brand: Nano Banana. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 50000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: auto, 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, 4:1, 1:4, 8:1, 1:8; editing geometry can follow the input image. - Reference image ceiling: 14; types: see workspace upload validation. - Output count: 1–4; format: png. ### Nano Banana 2 - Public model ID: nano-banana-2; provider brand: Nano Banana. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 20000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: auto, 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9; editing geometry can follow the input image. - Reference image ceiling: 14; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Nano Banana Pro - Public model ID: nano-banana-pro; provider brand: Nano Banana. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 20000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9; editing geometry can follow the input image. - Reference image ceiling: 14; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Nano Banana - Public model ID: nano-banana; provider brand: Nano Banana. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 20000 characters. - Resolutions: model default; default: model default. - Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Seedream 5 Lite - Public model ID: seedream-5-lite; provider brand: Seedream. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 5000 characters. - Resolutions: 2K, 3K; default: 2K. - Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Seedream 4.5 - Public model ID: seedream-4-5; provider brand: Seedream. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 3000 characters. - Resolutions: 2K, 4K; default: 2K. - Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Seedream V4 - Public model ID: seedream-v4; provider brand: Seedream. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 5000 characters. - Resolutions: 1K, 2K, 4K; default: 1K. - Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation. - Output count: 1–4; format: selected workspace/model default. ### Bria RMBG 2.0 - Public model ID: bria-background-remove; provider brand: Bria. - Modes: image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: optional/not required; maximum 0 characters. - Resolutions: original; default: original. - Aspect ratios: auto; editing geometry can follow the input image. - Reference image ceiling: 1; types: image/jpeg, image/png, image/webp; maximum 31457280 bytes per image. - Output count: 1; format: png. ### Illusion Diffusion - Public model ID: illusion-diffusion; provider brand: Illusion Diffusion. - Modes: image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 2000 characters. - Resolutions: 1K; default: 1K. - Aspect ratios: 1:1; editing geometry can follow the input image. - Reference image ceiling: 1; types: image/jpeg, image/png, image/webp; maximum 10485760 bytes per image. - Output count: 1; format: png. ### Z-Image - Public model ID: z-image; provider brand: Z-Image. - Modes: text. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 1000 characters. - Resolutions: model default; default: model default. - Aspect ratios: 1:1, 4:3, 3:4, 16:9, 9:16; editing geometry can follow the input image. - Reference image ceiling: 0; types: see workspace upload validation. - Output count: 1; format: selected workspace/model default. ### Nano Banana 2 Lite - Public model ID: nano-banana-2-lite; provider brand: Nano Banana. - Modes: text. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 50000 characters. - Resolutions: 1K; default: 1K. - Aspect ratios: auto, 1:1, 21:9, 16:9, 3:2, 4:3, 5:4, 4:5, 3:4, 2:3, 9:16, 4:1, 1:4, 8:1, 1:8; editing geometry can follow the input image. - Reference image ceiling: 0; types: see workspace upload validation. - Output count: 1–4; format: png. ### GPT Image 1.5 - Public model ID: gpt-image-1.5; provider brand: OpenAI. - Modes: text, image. Image editing needs reference image input; text mode does not require it. Layer decomposition needs one source image. - Prompt: required; maximum 32000 characters. - Resolutions: 1K; default: 1K. - Aspect ratios: 1:1, 3:2, 2:3; editing geometry can follow the input image. - Reference image ceiling: 10; types: see workspace upload validation. - Output count: 1–4; format: png. ## Video generation and processing ### MiniMax H3 Max Turbo - Public model ID: minimax-h3-max-turbo; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 50000 characters. - Resolution options: 480P, 768P, 1080P. - Duration options, seconds: 5–15, step 1. - Aspect ratios by mode: text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 480P, 5s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: true; seed control: true. ### MiniMax H3 - Public model ID: minimax-h3; provider brand: MiniMax; operation: generation. - Modes: reference, text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 7000 characters. - Resolution options: 2K. - Duration options, seconds: 5–15, step 1. - Aspect ratios by mode: reference: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: follows source. - Upload ceilings: 9 images, 3 videos, 3 audio files; selected mode validation still applies. - Defaults: reference, 2K, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: true; seed control: false. ### H3 Max - Public model ID: minimax-h3-max; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480P, 768P. - Duration options, seconds: 5–15, step 1. - Aspect ratios by mode: text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 768P, 5s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: true; seed control: true. ### Seedance 2.5 - Public model ID: seedance-2.5; provider brand: Seedance; operation: generation. - Modes: reference, text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480p, 720p. - Duration options, seconds: 4–30, step 1. - Aspect ratios by mode: reference: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; text: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; image: follows source. - Upload ceilings: 30 images, 10 videos, 10 audio files; selected mode validation still applies. - Defaults: reference, 480p, 4s, adaptive, audio on. - Audio control: optional. End-frame support: true; seed control: true. ### Seedance 2.0 Fast - Public model ID: seedance-2.0-fast; provider brand: Seedance; operation: generation. - Modes: reference, text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480p, 720p. - Duration options, seconds: 4–15, step 1. - Aspect ratios by mode: reference: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; text: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; image: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. - Upload ceilings: 9 images, 3 videos, 3 audio files; selected mode validation still applies. - Defaults: reference, 480p, 4s, adaptive, audio on. - Audio control: optional. End-frame support: true; seed control: false. ### Seedance 2.0 Pro - Public model ID: seedance-2.0-pro; provider brand: Seedance; operation: generation. - Modes: reference, text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480p, 720p, 1080p, 4k. - Duration options, seconds: 4–15, step 1. - Aspect ratios by mode: reference: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; text: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; image: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. - Upload ceilings: 9 images, 3 videos, 3 audio files; selected mode validation still applies. - Defaults: reference, 480p, 4s, adaptive, audio on. - Audio control: optional. End-frame support: true; seed control: false. ### Seedance 2.0 Mini - Public model ID: seedance-2.0-mini; provider brand: Seedance; operation: generation. - Modes: reference, text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480p, 720p. - Duration options, seconds: 4–15, step 1. - Aspect ratios by mode: reference: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; text: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; image: adaptive, 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. - Upload ceilings: 9 images, 3 videos, 3 audio files; selected mode validation still applies. - Defaults: reference, 480p, 4s, adaptive, audio on. - Audio control: optional. End-frame support: true; seed control: false. ### Seedance 1.5 Pro - Public model ID: seedance-1.5-pro; provider brand: Seedance; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 480p, 720p, 1080p. - Duration options, seconds: 4, 8, 12. - Aspect ratios by mode: text: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9; image: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 480p, 4s, 16:9, audio off. - Audio control: optional. End-frame support: true; seed control: true. ### Wan 2.7 - Public model ID: wan-2.7; provider brand: Wan; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 5000 characters. - Resolution options: 720p, 1080p. - Duration options, seconds: 2–15, step 1. - Aspect ratios by mode: text: 16:9, 9:16, 1:1, 4:3, 3:4; image: follows source. - Upload ceilings: 2 images, 1 videos, 1 audio files; selected mode validation still applies. - Defaults: text, 720p, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: true; seed control: false. ### Wan 2.6 - Public model ID: wan-2.6; provider brand: Wan; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 1500 characters. - Resolution options: 720p, 1080p. - Duration options, seconds: 5, 10, 15. - Aspect ratios by mode: text: follows source; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. ### Grok Imagine - Public model ID: grok-imagine; provider brand: Grok; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 5000 characters. - Resolution options: 720p. - Duration options, seconds: 6, 10. - Aspect ratios by mode: text: 16:9, 9:16, 1:1, 3:2, 2:3; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 6s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Styles: normal, fun, spicy. ### Kling 3.0 - Public model ID: kling-3.0; provider brand: Kling; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2500 characters. - Resolution options: 720p. - Duration options, seconds: 3–15, step 1. - Aspect ratios by mode: text: 16:9, 9:16, 1:1; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 3s, 16:9, audio off. - Audio control: optional. End-frame support: true; seed control: false. - Versions: standard, pro. ### PixVerse Extend - Public model ID: pixverse-extend; provider brand: PixVerse; operation: extend. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 360p, 540p, 720p, 1080p. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 0 images, 1 videos, 0 audio files; selected mode validation still applies. - Defaults: reference, 720p, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one MP4/MOV video under 30 seconds. Each extension adds 5 seconds; credits depend on resolution. ### ByteDance Video Upscaler - Public model ID: bytedance-video-upscaler; provider brand: ByteDance; operation: upscale. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: 1080p, 2k, 4k. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 0 images, 1 videos, 0 audio files; selected mode validation still applies. - Defaults: reference, 1080p, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one MP4/MOV video under 30 seconds. Credits depend on source duration and target resolution. ### Bria Video Background Removal - Public model ID: bria-video-background; provider brand: Bria; operation: remove-background. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: source. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 0 images, 1 videos, 0 audio files; selected mode validation still applies. - Defaults: reference, source, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one MP4/MOV video under 30 seconds and below 4000 × 4000 pixels. Download transparent WebM with original audio. ### PixVerse Person Swap - Public model ID: pixverse-swap; provider brand: PixVerse; operation: swap. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: 360p, 540p, 720p. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 1 images, 1 videos, 0 audio files; selected mode validation still applies. - Defaults: reference, 720p, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one portrait image and one MP4/MOV video up to 8 seconds. Videos longer than 5 seconds use the second price tier. ### Sync Lip Sync 2 - Public model ID: sync-lipsync-2; provider brand: Sync; operation: lipsync. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: source. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 0 images, 1 videos, 1 audio files; selected mode validation still applies. - Defaults: reference, source, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one MP4/MOV video and one WAV/M4A audio file, each under 30 seconds. Credits follow audio duration. ### Sync Talking Photo - Public model ID: sync-talking-photo; provider brand: Sync; operation: talking-photo. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: source. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 1 images, 0 videos, 1 audio files; selected mode validation still applies. - Defaults: reference, source, 5s, adaptive, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Input requirements: Upload one portrait image and one WAV/M4A audio file under 30 seconds. Credits follow audio duration. ### Hailuo 02 Pro - Public model ID: hailuo-02-pro; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 1080p. - Duration options, seconds: 6. - Aspect ratios by mode: text: follows source; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 6s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: true; seed control: false. ### Hailuo 02 Standard - Public model ID: hailuo-02-standard; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 768p. - Duration options, seconds: 6, 10. - Aspect ratios by mode: text: follows source; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 768p, 6s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: true; seed control: false. ### Hailuo 2.3 Pro - Public model ID: hailuo-2.3-pro; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 1080p. - Duration options, seconds: 6. - Aspect ratios by mode: text: follows source; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 6s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. ### Hailuo 2.3 Standard - Public model ID: hailuo-2.3-standard; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 768p. - Duration options, seconds: 6, 10. - Aspect ratios by mode: text: follows source; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 768p, 6s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. ### Seedance 1.0 Pro - Public model ID: seedance-1.0-pro; provider brand: Seedance; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2000 characters. - Resolution options: 480p, 720p, 1080p. - Duration options, seconds: 2–12, step 1. - Aspect ratios by mode: text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: true; seed control: true. ### Wan 2.5 - Public model ID: wan-2.5; provider brand: Wan; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 1500 characters. - Resolution options: 480p, 720p, 1080p. - Duration options, seconds: 5, 10. - Aspect ratios by mode: text: 16:9, 9:16, 1:1; image: follows source. - Upload ceilings: 1 images, 0 videos, 1 audio files; selected mode validation still applies. - Defaults: text, 1080p, 5s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: false; seed control: true. ### Kling 2.6 Pro - Public model ID: kling-2.6-pro; provider brand: Kling; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2500 characters. - Resolution options: 1080p. - Duration options, seconds: 5, 10. - Aspect ratios by mode: text: 16:9, 9:16, 1:1; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 5s, 16:9, audio on. - Audio control: optional. End-frame support: true; seed control: false. ### Veo 3.1 - Public model ID: veo-3.1; provider brand: Google; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 720p, 1080p, 4k. - Duration options, seconds: 4, 6, 8. - Aspect ratios by mode: text: 16:9, 9:16; image: auto, 16:9, 9:16. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 8s, 16:9, audio on. - Audio control: optional. End-frame support: false; seed control: true. ### HappyHorse 1.0 - Public model ID: happyhorse-1.0; provider brand: Alibaba; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2500 characters. - Resolution options: 720p, 1080p. - Duration options, seconds: 3–15, step 1. - Aspect ratios by mode: text: 16:9, 9:16, 1:1, 4:3, 3:4; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: true. ### HappyHorse 1.1 - Public model ID: happyhorse-1.1; provider brand: Alibaba; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 2500 characters. - Resolution options: 720p, 1080p. - Duration options, seconds: 3–15, step 1. - Aspect ratios by mode: text: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, 9:21, 5:4, 4:5; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 1080p, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: true. ### Grok Imagine 1.5 - Public model ID: grok-imagine-1.5; provider brand: Grok; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 4096 characters. - Resolution options: 480p, 720p, 1080p. - Duration options, seconds: 1–15, step 1. - Aspect ratios by mode: text: 16:9, 4:3, 3:2, 1:1, 2:3, 3:4, 9:16; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 6s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: false; seed control: false. ### H3 Max Styles - Public model ID: h3-max-styles; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 49868 characters. - Resolution options: 768P. - Duration options, seconds: 5–15, step 1. - Aspect ratios by mode: text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 768P, 5s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: false; seed control: true. - Styles: hand-drawn, vhs, retro-toon-70s, low-poly, 16bit-pixel. ### H3 Max Camera Control - Public model ID: h3-max-camera-control; provider brand: MiniMax; operation: generation. - Modes: image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 50000 characters. - Resolution options: 480P, 768P, 1080P. - Duration options, seconds: 3–15, step 1. - Aspect ratios by mode: image: follows source. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: image, 768P, 5s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: true. - Styles: orbit-right, orbit-left, push-in, pull-back, crane-up. ### H3 Max Lip Sync - Public model ID: h3-max-lip-sync; provider brand: MiniMax; operation: h3-lipsync. - Modes: reference. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: optional; maximum 2000 characters. - Resolution options: 480P, 768P, 1080P, 2K. - Duration options, seconds: 5. - Aspect ratios by mode: reference: follows source. - Upload ceilings: 1 images, 0 videos, 1 audio files; selected mode validation still applies. - Defaults: reference, 768P, 5s, adaptive, audio on. - Audio control: native audio, no off control. End-frame support: false; seed control: true. - Input requirements: Upload one portrait image and WAV/M4A audio lasting at least 5 seconds and under 30 seconds. Audio beyond 14.8 seconds is clipped. Credits follow the clipped audio duration. ### Gemini Omni - Public model ID: gemini-omni; provider brand: Google; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 20000 characters. - Resolution options: 720p. - Duration options, seconds: 3–10, step 1. - Aspect ratios by mode: text: 16:9, 9:16; image: 16:9, 9:16. - Upload ceilings: 1 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 8s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: false; seed control: false. ### H3 LoRA - Public model ID: h3-lora; provider brand: MiniMax; operation: generation. - Modes: text, image. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 50000 characters. - Resolution options: 480P, 768P, 2K, 4K. - Duration options, seconds: 5–15, step 1. - Aspect ratios by mode: text: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; image: follows source. - Upload ceilings: 2 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 768P, 5s, 16:9, audio on. - Audio control: native audio, no off control. End-frame support: true; seed control: true. ## Separate workflows ### Opus 5.5 · Motion Graphics - Public model ID: claude-opus-5-5-motion; provider brand: Anthropic; operation: generation. - Modes: text. Text uses a description; image uses a source image; reference mode requires the inputs selected in its workspace. These modes are not interchangeable. - Prompt: required; maximum 6000 characters. - Resolution options: 720p, 1080p. - Duration options, seconds: 10, 15, 20, 30. - Aspect ratios by mode: text: 16:9, 9:16, 1:1. - Upload ceilings: 4 images, 0 videos, 0 audio files; selected mode validation still applies. - Defaults: text, 720p, 20s, 16:9, audio off. - Audio control: no optional audio control configured. End-frame support: false; seed control: false. - Styles: sketch, motion. - Tool: https://nanobanana21ai.app/opus-5-5-video-generator. Separate structured scene and rendering workflow; no native soundtrack generation. Up to 4 approved image elements. The site renderer currently uses English display glyphs; written CJK/emoji layouts are not guaranteed. Review its operation quote, edit versions and render output in the workbench. - Humanize AI: https://nanobanana21ai.app/humanize-ai; maximum 12000 text characters; tones natural, friendly, professional. Improves clarity without a detector or factual-correctness guarantee. - Prompt Seen: https://nanobanana21ai.app/prompt-seen; one JPEG image under 1MB; optional instructions up to 1000 characters; styles detailed, concise, cinematic. Produces a reusable English prompt, not an identification of a private person or the source's original prompt. - GPT Image 2.5 extra settings follow its shared validation contract: quality auto, low, medium, high, xhigh, max, output format png, jpeg, webp, optional edit mask, size constraints recorded above. Auto quality/size quotes use the supported upper bound. - AI Playground https://nanobanana21ai.app/ai: demonstration cards for summarization, tagline writing, translation, image captioning, speech, image generation and editing. These are distinct examples with login/balance checks, not additional studio models or a promise that all upstream providers are available. Upload ceilings do not replace required-input validation. Text/image/reference modes, optional end frames, duration and audio constraints differ. Review generated media before using or publishing it.