Text2Movie

The film language each AI video model's docs let you write

Which models take a camera move from a reference?

2 of the 24 models: Moonvalley Marey and Seedance. The other 22 do not, in the docs read. As of 2026-10-13.

The register, split by the questionWhich models take a camera move from a reference? 2 models match and 22 do not.RegisterMatch2 models: Moonvalley Marey and SeedanceNo match22 models
Fig. 1 Counted from the models' docs on this page.
Models that answer the question, read 2026-10-13.
ModelCamera movesWhat the docs say
Moonvalley MareyDrawn or set from an imageCamera moves are made from a single image, and trajectories are drawn for elements
SeedanceFrom a referenceImages, audio and videos can serve as references for performance, light and camera movement

Inclusion rule. Models whose own docs meet the condition in the question. Order. Alphabetical by model name.

1Why it matters

Taking a move from a reference or a drawn path avoids describing it in words at all.

2From the docs

Create cinematic camera moves using just a single image.

Moonvalley, Marey

enables creators to transform an idea into visuals with full control over performance, lighting, shadow, and camera movement.

ByteDance Seed, Seedance 2.0

3The rest of the register

Not matching: Adobe Firefly, Google Veo, Grok Imagine, Higgsfield, HunyuanVideo 1.5, Kling AI, Leonardo Motion 2.0, LTX, LTX Studio, Luma Dream Machine, Midjourney, MiniMax Hailuo, Open-Sora 2.0, Pika, PixVerse, Runway, SceneMixer, SkyReels, Sora, Step-Video-TI2V, Vidu and Wan.

4Related questions