resolution

duration

Cinematic visuals with consistent motion and synced audio.

Text prompt for video generation. Works best with detailed, structured 'production brief' style prompts. Optional when a media/reference input is supplied.

Video aspect ratio. Set to 'adaptive' to let the model choose the best ratio based on inputs.

Video resolution. (Enum not documented on the model page.)

Generate synchronized audio with the video, including dialogue (use double quotes in prompt), sound effects, and background music.

Random seed. Set for reproducible generation (reproducibility not guaranteed).

seedance 2.5

Cinematic results with multiple references and free audio.

The text prompt. Put dialogue in double quotes for lip-sync.

The aspect ratio of the generated video. 'adaptive' lets the model choose.

The duration of the generated video in seconds. -1 lets the model intelligently pick the length.

Whether to generate synchronized dialogue, SFX, and music.

seedance 2.0

Grok Imagine 1.5

Cinematic quality, sound and visuals in sync.

Alibaba Happy Horse

Quick animations and product demos, cheaper.

The aspect ratio of the generated video frame.

The duration of the generated video in seconds.

grok imagine

Cinematic quality with smooth continuity.

Text prompt describing the video. Include scene, camera movement, lighting, mood, and audio for best results. Put anything to avoid in the prompt too (e.g. 'no dialogue').

Resolution of the generated video. 360p is a fast draft mode.

Reference images that guide the subject and style of the video (not used as literal frames). Use to feature specific characters or objects.

gemini omni 1.1

referenceVideoKeys

Video duration in seconds. Set to -1 for intelligent duration (the model picks the best length).

Reference videos (up to 10, combined duration ≤30s) for motion transfer, style, editing, and extension. Cite in prompt as [Video1].

Reference audio files (up to 10, combined duration ≤30s) for audio-driven generation and lip-sync. Requires at least one reference image or video. Cite in prompt as [Audio1].

refVideoCost

Reference videos for motion transfer/style/editing, cited as [Video1]. Up to 3, total duration <= 15s.

Reference audios for audio-driven generation / lip-sync, cited as [Audio1]. Up to 3, total <= 15s. Requires at least one reference image or video.

generateAudio

Keep your character's look and voice identical across scenes.

The text prompt. Supports template references such as <<<image_1>>> and <<<video_1>>>.

Reference images for elements/scenes/styles, cited as <<<image_N>>>. Maximum 7 (or 4 when a reference video is also used).

Reference video for editing or style/camera reference. Duration 3-10s, 720-2160px per side, max 200MB.

How the reference video is used. 'feature' = style/camera reference, 'base' = video editing.

Whether to generate native audio. Mutually exclusive with reference_video.

Resolution tier. standard = 720p, pro = 1080p, 4k = 4K. 4K does not support reference_video.

The aspect ratio of the generated video. Required when not using a start frame or video editing.

The duration of the generated video in seconds. Ignored for video editing (base).

Multi-shot JSON array. Maximum 6 shots, minimum 1 second per shot; total must equal the duration.

kling v3 omni pro

The text prompt to guide video generation.

The aspect ratio of the generated video. Ignored when a start image is provided.

Resolution tier. standard = 720p, pro = 1080p, 4k = 4K.

Whether to generate native audio (dialogue, SFX, ambient) with lip sync.

Multi-shot JSON array, e.g. [{"prompt":"...","duration":N}]. Maximum 6 shots, minimum 1 second per shot; total must equal the duration.

kling v3 pro

Whether to generate native audio for the video.

ltx 2.3 pro

Faster and cheaper choice for quality videos with sound.

Input image to start generating from. Ideal images are 16:9 or 9:16 and 1280x720 or 720x1280, depending on the aspect ratio you choose.

Description of what to exclude from the generated video

veo 3.1 fast

Realistic videos with precise motion control and accurate lip-sync.

The text prompt to guide video generation. Mutually exclusive with multi_prompt.

Multi-shot prompt list. Each element has a prompt and a duration. Overrides the single prompt.

The CFG (Classifier Free Guidance) scale is a measure of how close you want the model to stick to your prompt.

Whether to generate native audio for the video. Supports Chinese and English voice output; other languages are automatically translated to English.

Shot handling. 'intelligent' lets the model auto-determine shots.

kling v3 standard

The text prompt to generate a video from. Should read like a video caption, not a command; negating language (e.g. 'no cats') is not supported.

Width and height of the output video. Only 1280x720 (16:9) is currently supported.

Duration of the output video in seconds. Single-shot (TEXT_VIDEO) is fixed at 6.

Frame rate of the output video. 24 is the only supported value.

Initial noise seed. Same seed + same parameters reproduces a result; change it for a new variation.

Amazon Nova Reel 1.0

The text prompt to generate the video from.

The aspect ratio of the generated video frame

The duration of the generated video in seconds

Specify elements or qualities to avoid in the generated video, such as 'blur', 'distort', or 'low quality'.

Kling Video v1.6 standard

High quality videos with realistic people and natural lip-sync.

veo 3.1

Video aspect ratio. Ignored when an image is used.

Video resolution - 480p faster, 720p balanced, 1080p higher quality.

Random seed. Set for reproducible generation.

ByteDance Seedance 1.0 Pro Fast

audioCharge

Cinematic videos with expressive motion and synced audio.

Generate audio synchronized with the video.

ByteDance Seedance 1.5 Pro

Cinematic videos with built-in audio at the lowest rates for this quality.

The text prompt to generate a video from.

The CFG (Classifier Free Guidance) scale is a measure of how close you wantthe model to stick to your prompt.

Whether to generate native audio for the video. Supports Chinese and English voice output. Other languages are automatically translated to English. For English speech, use lowercase letters; for acronyms or proper nouns, use uppercase.

Kling Video v2.6

High-end motion quality for your cinematic visuals.

The CFG (Classifier Free Guidance) scale is a measure of how close you want
            the model to stick to your prompt.