Video models: six things the docs say about the camera

Twenty-four AI video models and tools, each read in its own documentation on six points: how a camera move is asked for, whether shot sizes are defined, whether one generation can hold several shots, start and end frames, negative prompts, and a setting for how much the picture moves. As of 2026-09-25.

Fig. 1 The register's columns fall into these two halves.

Fig. 2 Lit: the docs cover it. Dark: they do not, or could not be read.

Camera readings by model, read 2026-09-25.
Model	Camera moves	Framing	Cuts	Frames	Negative prompt	Motion amount
Adobe Firefly	Commands or presets	Options to pick from	None in the docs	First and last frame	None in the docs	None in the docs
Google Veo	Words in the prompt	Defined in a glossary	Editing terms only	First and last frame	A negative prompt field	None in the docs
Grok Imagine	Words in the prompt	None in the docs	None in the docs	First frame only	None in the docs	None in the docs
Higgsfield	Commands or presets	None in the docs	None in the docs	None in the docs	None in the docs	None in the docs
HunyuanVideo 1.5	Words in the prompt	Shown in examples	None in the docs	First frame only	A negative prompt field	None in the docs
Kling AI	Words in the prompt	Shown in examples	Several shots per clip	First and last frame	None in the docs	None in the docs
Leonardo Motion 2.0	Commands or presets	Options to pick from	None in the docs	First frame only	A negative prompt field	A motion setting
LTX	Words in the prompt	Shown in examples	Several shots per clip	Several keyframes	None in the docs	None in the docs
LTX Studio	Commands or presets	Options to pick from	None in the docs	First frame only	None in the docs	None in the docs
Luma Dream Machine	Commands or presets	None in the docs	None in the docs	First and last frame	None in the docs	None in the docs
Midjourney	None in the docs	None in the docs	None in the docs	First and last frame	None in the docs	A motion setting
MiniMax Hailuo	Commands or presets	None in the docs	None in the docs	First and last frame	None in the docs	None in the docs
Moonvalley Marey	Drawn or set from an image	None in the docs	None in the docs	Several keyframes	None in the docs	None in the docs
Open-Sora 2.0	None in the docs	Shown in examples	None in the docs	First frame only	None in the docs	A motion setting
Pika	Words in the prompt	None in the docs	None in the docs	First and last frame	None in the docs	None in the docs
PixVerse	Commands or presets	None in the docs	None in the docs	First and last frame	None in the docs	A motion setting
Runway	Words in the prompt	None in the docs	None in the docs	Several keyframes	Advised against	None in the docs
SceneMixer	Words in the prompt	Shown in examples	Several shots per clip	None in the docs	Advised against	None in the docs
Seedance	From a reference	None in the docs	Several shots per clip	N/A, docs unreadable	N/A, docs unreadable	N/A, docs unreadable
SkyReels	Words in the prompt	Shown in examples	None in the docs	First and last frame	None in the docs	None in the docs
Sora	Words in the prompt	Shown in examples	Several shots per clip	First frame only	None in the docs	None in the docs
Step-Video-TI2V	Words in the prompt	None in the docs	None in the docs	First frame only	None in the docs	A motion setting
Vidu	None in the docs	None in the docs	None in the docs	First and last frame	None in the docs	A motion setting
Wan	Words in the prompt	Defined in a glossary	Several shots per clip	First and last frame	A negative prompt field	None in the docs

Inclusion rule. AI video models and tools whose makers publish a prompt guide, API reference or product page that says something about camera, framing, cuts or frames. Order. Alphabetical by model name.

1How to read a row

The first three columns are about writing the shot: moves, framing and cuts. The last three are the controls around the prompt: frames that pin the start and end, a negative prompt, and a setting for how much moves. A model can be strong on one side and silent on the other; the docs decide, not the model's output.

None in the docs means the pages read say nothing about it. It does not mean the model cannot do it, only that its maker has not written it down where a user would look.

2Every model

Adobe Firefly — shot size, angle and motion menus
Google Veo — a full shot glossary
Grok Imagine — video from text or a still
Higgsfield — more than fifty motion presets
HunyuanVideo 1.5 — camera moves written into demo prompts
Kling AI — a multi-shot switch and shot list format
Leonardo Motion 2.0 — Motion Control presets with a weight
LTX — named cuts and several keyframes
LTX Studio — camera presets and a shot editor
Luma Dream Machine — camera words from a published list
Midjourney — low or high motion
MiniMax Hailuo — fifteen bracketed camera commands
Moonvalley Marey — camera moves from a single image
Open-Sora 2.0 — a motion score in the prompt
Pika — keywords and Pikaframes
PixVerse — a camera_movement field
Runway — camera in words, keyframes in an app
SceneMixer — numbered shots inside each segment
Seedance — native multi-shot storytelling
SkyReels — start and end frame control
Sora — a shot list in one clip
Step-Video-TI2V — a supported-moves table and motion score
Vidu — a movement amplitude setting
Wan — a prompt dictionary and multi-shot formula

3Other reels

Camera words across every model are under camera words; questions such as which models take an end frame are under shot questions; how the docs were read is on the method page.