# Generation models

Every image, video, music, sound-effect and voice model a Block can use and the settings each accepts, generated from the SDK catalog.

## How models are used
Capability Blocks (`image.generate`, `video.generate`, `audio.music`, `audio.sfx`, `audio.speech`) and JavaScript Blocks (the `media.generate` action) can use every model below, the same catalog the app's own Blocks use. Users pay with StillMade credits at StillMade's price; a Block declares default settings and the user can change them before running. A setting a model has no knob for is `"default"`.

## Image models
| Model | Name | Aspect ratios | Resolutions | Qualities | Reference images | Seed |
| --- | --- | --- | --- | --- | --- | --- |
| `nano-banana-2` | Nano Banana 2 | auto, 21:9, 16:9, 3:2, 4:3, 1:1, 4:5, 3:4, 2:3, 9:16 | 1K, 2K, 4K | model default | up to 8 | yes |
| `nano-banana-2-lite` | Nano Banana 2 Lite | auto, 21:9, 16:9, 3:2, 4:3, 1:1, 4:5, 3:4, 2:3, 9:16 | 1K, 2K, 4K | model default | up to 8 | yes |
| `nano-banana-pro` | Nano Banana Pro | auto, 21:9, 16:9, 3:2, 4:3, 1:1, 4:5, 3:4, 2:3, 9:16 | 1K, 2K, 4K | model default | up to 8 | yes |
| `gpt-image-2` | GPT Image 2 | auto, 16:9, 9:16, 1:1, 4:3, 3:4 | 1K, 2K, 4K | low, medium, high | up to 8 | no |
| `gpt-image-1.5` | GPT-Image 1.5 | 16:9, 9:16, 1:1 | model default | low, medium, high | no | no |
| `kling-image-v3` | Kling Image V3 | 21:9, 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3 | 1K, 2K | model default | no | no |
| `seedream-v4.5` | Seedream 4.5 | 16:9, 9:16, 1:1, 4:3, 3:4 | 2K, 4K | model default | no | yes |

## Video models
| Model | Name | Seconds | Resolutions | Aspect ratios | First frame | Last frame | Native sound |
| --- | --- | --- | --- | --- | --- | --- | --- |
| `seedance-2.0` | Seedance 2.0 | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 | 480p, 720p, 1080p, 4k | auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16 | yes | yes | yes |
| `seedance-2.0-fast` | Seedance 2.0 Fast | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 | 480p, 720p | auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16 | yes | yes | yes |
| `seedance-2.0-mini` | Seedance 2.0 Mini | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 | 480p, 720p | auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16 | yes | yes | yes |
| `kling-v3-turbo` | Kling 3.0 Turbo | 5, 10 | model default | 16:9, 9:16, 1:1 | yes | no | no |
| `wan-2.6-flash` | Wan 2.6 Flash | 3, 4, 5 | model default | 16:9, 9:16, 1:1 | yes | no | no |
| `veo-3.1-lite` | Veo 3.1 Lite | 4, 6, 8 | 720p, 1080p | 16:9, 9:16 | yes | yes | yes |
| `vidu-q3-turbo` | Vidu Q3 Turbo | 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 | model default | model default | yes | yes | yes |
| `wan-2.7` | Wan 2.7 | 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 | 720p, 1080p | model default | required | yes | no |

Veo 3.1 Lite at 1080p is always 8 seconds. Generated images are stored as measured PNGs (up to 8192 pixels a side); videos are stored as the provider's MP4.

## Music and sound effects
| Operation | Model | Length | Price |
| --- | --- | --- | --- |
| `audio.music` | `ace-step` | 5 to 240 whole seconds (suggested default 30) | The app's music price per started minute |
| `audio.sfx` | `sound-effects` | 0.5 to 22 seconds in tenths (suggested default 4) | The app's sound-effect price per clip |

## Voices
`audio.speech` Blocks can suggest any of these voice ids in `settings.voice`. A server offers the voices whose provider it has configured. Price follows the app's Voiceover rate for the text's length.

| Voice id | Name | Provider |
| --- | --- | --- |
| `asteria` | Asteria · Clear, confident female | deepgram |
| `adam` | Adam · Dominant, firm male | elevenlabs |
| `liam-social` | Liam · Social · Energetic — reels & shorts | elevenlabs |
| `onyx` | Onyx · Deep, steady male narrator | openai |
| `nova` | Nova · Bright, clear female narrator | openai |
| `luna` | Luna · Friendly, youthful female | deepgram |
| `odysseus` | Odysseus · Calm, smooth, professional male | deepgram |
| `aurora` | Aurora · Cheerful, expressive female | deepgram |
| `apollo` | Apollo · Confident, comfortable male | deepgram |
| `thalia` | Thalia · Confident, energetic female | deepgram |
| `atlas` | Atlas · Enthusiastic, confident male | deepgram |
| `hera` | Hera · Warm, professional female | deepgram |
| `arcas` | Arcas · Natural, smooth male | deepgram |
| `cora` | Cora · Smooth, melodic female | deepgram |
| `zeus` | Zeus · Deep, trustworthy male | deepgram |
| `orion` | Orion · Approachable, calm male | deepgram |

Music, sound effects and catalog voices are stored as the provider's measured MP3 or WAV.