Hailuo 2.3 video model

Use Hailuo 2.3 with its current video limits visible.

Vizify currently lists 2 related policies: Hailuo 2.3, Hailuo 2.3 Pro. This page shows the verified public parameter envelope they share and the workflows that accept it.

2 available policies Duration: 6/10s Output: 768p / 1080p 1 reference image required Silent video Image start required
Verified catalog facts

What Vizify exposes for this model family.

These facts are generated from the current public catalog, not copied from provider marketing.

Variants2

Hailuo 2.3 · Hailuo 2.3 Pro

Inputs2

Text · Image

Outputs1

Video

AccessTask-based

Video generation

StatusAvailable

Reviewed 2026-08-02

Best for

Where this family fits.

Preparing a bounded Hailuo 2.3 shot brief

Checking 6/10s, 768p / 1080p and auto before you run

Deciding between the 6/10s Hailuo 2.3 lengths before writing the prompt

Cuts you will score or narrate yourself, because Hailuo 2.3 returns silent video

Animating the one reference image Hailuo 2.3 requires

Current limits

What not to assume.

Duration, resolution, audio and reference support differ across Hailuo 2.3, Hailuo 2.3 Pro, so the figures above are the family union rather than one policy.

Hailuo 2.3 returns video without an audio track, so any sound has to be added after the render.

Motion can drift in identity, geometry, small text, logos and continuity, so Hailuo 2.3 output still needs review before publishing.

A Hailuo 2.3 request is asynchronous: a pending job is not a finished clip.

2 resolution-and-duration combinations are refused on purpose, so 1080p is not available at every Hailuo 2.3 length.

What the Hailuo 2.3 envelope allows

Hailuo 2.3 currently lists 6/10s durations, 768p / 1080p output and auto aspect ratios. Length is a choice between 6/10s rather than a free range, so the cut is decided before the prompt is written. 2 resolution-and-duration pairs are rejected, so 1080p is not available at every length.

Choosing the workflow

Hailuo 2.3 starts from a reference image only, so the source frame decides framing, subject and palette before the prompt is read. Exactly one reference frame is required per request. The clip arrives silent, so the audio pass belongs to the edit.

How a request runs

Vizify inspects the selected policy before it sends the request, so a value outside the published envelope is refused rather than quietly adjusted. Video generation is asynchronous: a pending job is not a completed artifact, and the returned clip still needs a review pass for identity, geometry, small text and continuity before it is published.

Evidence boundary

Availability and parameters were checked against Vizify’s generation catalog on 2026-08-02. These product facts do not establish market ranking, external pricing, or benchmark leadership.

FAQ

Questions about Hailuo 2.3 video models in Vizify

Is Hailuo 2.3 available in Vizify?

Yes. Hailuo 2.3, Hailuo 2.3 Pro are listed in the current public video catalog as 2 selectable policies.

Which Hailuo 2.3 variants does this page cover?

It covers Hailuo 2.3, Hailuo 2.3 Pro. Each request uses one selected policy, and duration, resolution and audio can differ between them.

What duration, resolution and aspect ratios does Hailuo 2.3 support?

The catalog lists 6/10s, 768p / 1080p and auto. Only the listed lengths are accepted, so the cut is chosen before the prompt is written.

Does Hailuo 2.3 generate audio?

No. Hailuo 2.3 returns silent video, so music, dialogue and effects are added in a later step.

How many reference images can Hailuo 2.3 use?

Hailuo 2.3 is image-to-video only: exactly one reference image is required for every request.

Where do I start a Hailuo 2.3 video?

Open a compatible Vizify video tool, prepare the prompt and any required reference image, then let Vizify inspect the selected Hailuo 2.3 policy before execution.