Hailuo AI Video Generator

Hailuo is MiniMax’s video model family. Hailuo 2.3, listed as a legacy model on MiniMax’s API, turns a prompt or a first frame into a 6 or 10 second clip at 768P or 1080P, at 24 fps. The limits below are from MiniMax’s video generation docs.

01 — Capabilities

A short clip, steered with camera tags

Hailuo 2.3 is specified as text-to-video or image-to-video. Length and resolution are a pair: 10 seconds is not available at 1080P.

01
Only 6 or 10 seconds
Those are the durations MiniMax documents for Hailuo 2.3. 1080P is 6 seconds only. 10 seconds stays at 768P. There is no 15-second or 30-second mode on this line.
02
Fifteen camera commands in brackets
The prompt can carry tags such as [Push in], [Pan left] or [Tracking shot]. Combined moves go in one pair of brackets, three at most. Natural language also works, but the docs say the tags are more accurate.
03
A first frame on 2.3; a last frame on Hailuo 02
Hailuo 2.3 animates one still as the opening image. First-and-last-frame interpolation is documented on Hailuo 02, not on 2.3. Hailuo 2.3 Fast is image-to-video only.
04
No soundtrack in the Hailuo API
The video generation request has no audio switch. Speech and music are separate MiniMax products. Treat a Hailuo clip as picture unless you add sound somewhere else.

02 — Prompt to output

Prompts that match the clips above

Each prompt describes the picture on its card. “Use this prompt” sends that text to Create. The generator on this page runs Seedance 2.5, not Hailuo.

01 · Portrait
PromptClose-up of a young man with curly hair in a grey knit sweater, soft window light. He glances away, then smiles. Slow push-in, shallow depth of field.
02 · Product
PromptA white leather sneaker on a pale grey studio backdrop, soft diffused light. The camera orbits slowly. Stitching stays sharp, the look stays minimal.
03 · Street
PromptA slow dolly down a cobblestone lane in an old European village. Climbing roses on stone walls, café tables in morning sun, a few people walking in the distance.
04 · Character
PromptA young woman with shoulder-length brown hair in an olive t-shirt, late-afternoon gold light. She turns her head and looks off. Breeze in her hair, a handheld feel.

03 — Where it fits

Jobs that fit a 10-second ceiling

Hailuo 2.3’s spec is one short shot with camera control. Longer stories have to be cut together somewhere else, and they still have no native audio.

One named camera move
A push-in, a pan or a tracking shot can be written as a tag. That is the control MiniMax documents. A multi-location sequence does not fit in 6 or 10 seconds.
You already have the opening still
Image-to-video takes one first frame. Aspect follows that image. Do not expect Hailuo 2.3 to also lock a last frame; that path is Hailuo 02.
A short beat of physical action
MiniMax positions Hailuo 2.3 around instruction following and physics. A toss, a turn or a landing can sit inside 6 seconds at 1080P.
Draft at 768P, then pay for 1080P
10-second tests stay at 768P. 1080P is the 6-second setting. The docs do not offer 4K on this family.

04 — Verdict

What the Hailuo API actually lists

Strengths

Text-to-video and image-to-video on Hailuo 2.3
Fifteen documented camera commands in the prompt
768P at 6 or 10 seconds; 1080P at 6 seconds, 24 fps
Prompt optimizer can be turned off for tighter control

Trade-offs

No native audio on the Hailuo video API
10 seconds is 768P only; 1080P is 6 seconds only
Last-frame interpolation is Hailuo 02, not Hailuo 2.3
Hailuo is not available on ImgToVid

05 — Side by side

Hailuo 2.3, Veo 3.1 and Kling 4.0

Length, sound and steering are where these three stop being interchangeable. MiniMax now lists Hailuo 2.3 as a legacy video model next to H3.

	Hailuo 2.3	Veo 3.1	Kling 4.0
Developer	MiniMax	Google DeepMind	Kuaishou
Single clip	6 or 10 seconds	4, 6 or 8 seconds	3–30 seconds
Picture	768P or 1080P at 24 fps	720p, 1080p, 4K at 24 fps	720p, 1080p and 4K listed
Sound	Not in the Hailuo video API	Native, with the picture	Stereo, with lip sync
Steering	Camera tags, first frame	First frame, last frame, up to 3 references	Up to 15 references, or up to 10 keyframes
On ImgToVid	Coming soon	Coming soon	Coming soon

06 — How it works

Start a clip on ImgToVid

Step 1
Describe one shot
Subject, action, camera. If you wanted Hailuo’s bracket tags, write the move in plain language instead. Hailuo will not be the model that runs it.
Step 2
Use the generator on this page
It runs Seedance 2.5, because Hailuo is not offered here. The prompt and the settings carry into Create.
Step 3
Generate in Create
Membership and credits are checked there. Coming back to this page does not switch the model over to Hailuo.

07 — FAQ

Hailuo, in plain terms

What is Hailuo?

Hailuo is MiniMax’s video model family. The current documented Hailuo line is Hailuo 2.3, with a Fast image-to-video variant and the earlier Hailuo 02. MiniMax’s model map now lists these as legacy next to MiniMax H3.

How long is a Hailuo 2.3 clip?

6 or 10 seconds. 1080P is 6 seconds only. 10 seconds is documented at 768P. Default duration is 6 seconds.

Does Hailuo generate audio?

The Hailuo video generation API does not document an audio switch or native soundtrack. MiniMax speech and music are separate products.

Can I lock a first and last frame?

Hailuo 2.3 takes a first-frame image for image-to-video. First-and-last-frame interpolation is documented on Hailuo 02. Hailuo 2.3 Fast requires a first frame and does not list text-only generation.

Can I generate Hailuo videos on ImgToVid?

No. The generator on this page runs Seedance 2.5 so you can draft a shot. Hailuo itself is marked coming soon.

Can I use the results commercially?

Commercial use of anything you generate on ImgToVid depends on your plan. Check Pricing and our Terms before you publish, and make sure prompts and uploads do not infringe someone else’s rights.