Video Models

The best AI video generation models — Seedance 2.5, Kling 3, Veo 3.1, Wan 3.0, Hailuo, LTX 2.5 and Flux 3 — with credits per clip shown before you run.

Explore by category

Models

Seedance 2.595
Seedance 2.5
Seedance 2.5 is ByteDance's latest AI video model — native 30-second clips and up to 50 multimodal references.
31
Seedance 2
Seedance 2.0 is ByteDance's multimodal AI video model: cinematic motion, sound and up to 15-second clips.
34
Flux 3
Create videos up to 20 seconds in one generation, with sound built in.
Minimax H326
Minimax H3
MiniMax H3 (Hailuo 03) is a frontier video model that generates 2K video up to 15 seconds long
16
H3 Max
fal's post-trained MiniMax H3, made with MiniMax's H3 team and tuned for prompt adherence and speed. Text or image to video at 480P or 768P.
H3 Max Turbo
8
H3 Max Turbo
A post-trained H3 Max variant tuned for stronger prompt adherence and better aesthetics, co-optimized with fal's inference stack for speed.
Gemini Omni32
Gemini Omni
Latest video generation and editing model from Google
Wan 3.040
Wan 3.0
Wan 3.0 is Alibaba's next-generation video model. Up to 30 seconds of 1080p video in a single pass, with audio baked into the same generation.
Wan 3.0 Prime
56
Wan 3.0 Prime
Wan 3.0 Prime — faster generation with fluid motion and strong scene fidelity, and a reference mode that keeps character identity across images, video and audio references.
LTX 2.532
LTX 2.5
Generate up to 10 seconds of 1080p video with sync'd audio — pro quality or fast drafts.
Kling 334
Kling 3
Kling 3 Pro — cinematic motion, native audio and multi-shot prompting up to 15 seconds.
Veo 3.148
Veo 3.1
Google's Veo 3.1 — lifelike video with synchronized dialogue and sound effects.
Grok Imagine Video 1.534
Grok Imagine Video 1.5
xAI's video model — snappy, expressive clips from text or a starting image.
Luma Ray 3.26
Luma Ray 3.2
Luma's Ray 3.2 — HDR-ready video with reasoning-driven prompt understanding.
PixVerse V612
PixVerse V6
Stylized, effect-heavy video generation for social clips and transitions.
Vidu Q331
Vidu Q3
Vidu Q3 — character-consistent video generation with reference subjects.