Skip to main content
Use these models with the video generation endpoints. Reference a model by its id in the model field.
The exact set your key can call, and each model’s supported resolutions, durations, and options, is returned by GET /v1/models. Video is priced per second of output; for per-model rates see deepshi.ai.
Recently added models appear first.

Gemini Omni Flash 1.1

gemini-omni-flash-1.1 · Prompt 20KThe upgraded Flash tier, from 360p up to 4K, with reference-guided generation from images and short videos. Image to video   Audio

Grok Imagine Video 1.5

grok-imagine-video-1.5 · Prompt 4KxAI’s newest video tier, adding 1080p and reference images, with audio generated automatically. Image to video   Audio

LTX 2.5

ltx-2.5 · Prompt 5KLightricks’ flagship audio-video tier at 720p and 1080p, with camera-motion and frame-rate control. Image to video   Audio

LTX 2.5 Fast

ltx-2.5-fast · Prompt 5KThe speed-optimized LTX 2.5 tier, up to 20 seconds and 4K, with native audio. Image to video   Audio

MiniMax H3 Max

minimax-h3-max · Prompt 50KThe top Hailuo 03 tier at 480P and 768P, with an optional end frame to pin first and last images. Image to video

Wan 3.0

wan-3.0 · Prompt 5KAlibaba’s flagship Wan, 2 to 30 second clips up to 1080p with audio on by default and first/last-frame control. Image to video   Audio

Wan 3.0 Prime

wan-3.0-prime · Prompt 5KThe premium Wan 3.0 tier, adding an optional deliberate thinking pass and reference-guided generation. Image to video   Audio

FLUX.3

flux-3 · Prompt 5KBlack Forest Labs’ audio-video model, 5 to 20 second clips at 720p or 1080p with native audio. Image to video   Audio

Seedance 2.5

seedance-2.5 · Prompt 5KByteDance’s Seedance 2.5, native single-shot clips up to 30 seconds at up to 1080p, with audio. Image to video   Audio

Kling O3 4K

kling-o3-4k · Prompt 2.5KThe Kling O3 tier that outputs 4K in a single pass, keeping multi-shot storyboarding and character consistency. Image to video   Audio

Kling O3

kling-o3 · Prompt 2.5KThe entry O3 tier, with the line’s multi-shot storytelling at a lower per-second rate. Image to video   Audio

Kling v3 Turbo Pro

kling-v3-turbo-pro · Prompt 2.5KThe 1080p tier of Kling 3.0 Turbo, with improved lipsync and multi-shot generation. Audio is always on. Image to video   Audio

Kling v3 Turbo

kling-v3-turbo · Prompt 2.5KThe fast Kling 3.0 tier at 720p, with native always-on audio at a flat per-second rate. Image to video   Audio

MiniMax H3

minimax-h3 · Prompt 2KHailuo 03, 5 to 15 second clips from 480P up to 4K, from text, an image, or up to 12 reference files. Image to video

Seedance 2.0 Mini

seedance-2.0-mini · Prompt 5KThe smallest Seedance tier at 480p and 720p, with audio and flexible durations. Image to video   Audio

Veo 3.1 Lite

veo3.1-lite · Prompt 20KThe entry Veo 3.1 tier at 720p and 1080p with audio, at the lowest Veo price. Image to video   Audio

Kling v3 4K

kling-v3-4k · Prompt 2.5KThe Kling v3 tier that renders natively at 4K, with audio and voice control. Image to video   Audio + voice

Happy Horse 1.1

happy-horse-1.1 · Prompt 2.5KAlibaba’s Happy Horse, at 720p and 1080p with flexible clip durations. Image to video

LTX-2.3

ltx-2.3 · Prompt 5KLightricks’ open model, up to 4K at high frame rates, with audio. Image to video   Audio

PixVerse V6

pixverse-v6 · Prompt 5KPixVerse’s fast generator from 360p to 1080p, for short clips. Image to video

Gemini Omni Flash

gemini-omni-flash · Prompt 20KGoogle’s Gemini Omni Flash at 720p, with audio built in. Image to video   Audio

Wan 2.5

wan-2.5 · Prompt 1.5KText-to-video and image-to-video at up to 1080p. Image to video

Wan 2.6

wan-2.6 · Prompt 1.5KA newer Wan tier at 720p and 1080p, with longer clip lengths. Image to video

Wan 2.7

wan-2.7 · Prompt 5KThe latest Wan tier, with flexible clip durations at 720p and 1080p. Image to video

Veo 3.1

veo3.1 · Prompt 20KGoogle’s flagship generator, up to 4K with optional generated audio. Image to video   Audio

Veo 3.1 Fast

veo3.1-fast · Prompt 20KThe faster, lower-cost Veo 3.1 tier, same resolutions and audio. Image to video   Audio

Sora 2

sora-2 · Prompt 5KOpenAI’s Sora 2 at 720p, with audio generated as part of the clip. Image to video   Audio

Sora 2 Pro

sora-2-pro · Prompt 5KThe higher-fidelity Sora 2 tier, up to true 1080p, with native audio. Image to video   Audio

Kling v3

kling-v3 · Prompt 2.5KKuaishou’s Kling v3, with optional audio and voice control. Image to video   Audio + voice

Kling v3 Pro

kling-v3-pro · Prompt 2.5KThe pro Kling v3 tier, for higher-fidelity motion, with audio and voice control. Image to video   Audio + voice

Kling o3 Pro

kling-o3-pro · Prompt 2.5KA Kling tier tuned for prompt adherence, with optional audio. Image to video   Audio

Grok Imagine Video

grok-imagine-video · Prompt 4KxAI’s fast video generator at 480p and 720p, with many aspect ratios. Image to video

Seedance 2.0

seedance-2.0 · Prompt 5KByteDance’s Seedance, up to 4K with optional audio and flexible durations. Image to video   Audio

Seedance 2.0 Fast

seedance-2.0-fast · Prompt 5KThe faster, lower-cost Seedance tier at 480p and 720p. Image to video   Audio
Not sure where to start? grok-imagine-video is a fast, economical default, veo3.1 is a strong all-round choice with audio, and sora-2 is a good option when you want sound built in.