Hailuo 02: Text & Image to Video Online

Text & Image to Video4.0 / 5Flexible content filter

Bottom line

MiniMax's Hailuo 02 leans on physics realism and camera control, from text or image, in 6 or 10 second clips. Start from a prompt or a still image in the Text to Video tool; VdoBloom shows the exact credit cost before each Hailuo 02 run, with no separate Hailuo account.

Example output generated with Hailuo 02

Physics is where AI video usually falls apart — liquids that defy gravity, hair that clips through shoulders, objects that collide without consequence. Hailuo 02, from Chinese AI lab MiniMax, stakes its reputation on getting exactly that right. The model is positioned around next-generation physics simulation and camera control, which shows up in shots involving weight, momentum, and interaction: things falling, splashing, swinging, or being thrown tend to move like they should.

It works from either a text prompt or a starting image, so you can invent a scene from scratch or animate an existing frame. Clips come in 6 or 10 second lengths at 768P or 1080P, and generation costs VdoBloom credits with the price shown up front.

One more thing sets it apart in this catalog: its HOT content label. Where the Kling and Hailuo 2.3 models run strict filters, Hailuo 02's filtering is more permissive, giving creators working on stylized action, horror atmospheres, or edgier concepts more room before a prompt gets refused. It carries a 4.0 rating.

What creators use Hailuo 02 for

  • Generating action shots with believable weight and momentum — a skateboarder landing a trick, water crashing over rocks, cloth caught in wind.
  • Directing deliberate camera moves like push-ins and orbits around a subject, using the model's camera-control strengths.
  • Producing moodier or more intense concepts, such as horror atmospheres or stylized combat, that stricter models tend to refuse.

How to use Hailuo 02 on VdoBloom

  1. 1Open the Text to Video tool and select Hailuo 02 from the model picker.
  2. 2Write your prompt — or upload a starting image for image-to-video.
  3. 3Choose your duration (6–10s), quality (768P/1080P), review the credit cost, and generate.

Frequently asked questions

How strict is Hailuo 02's content filter?

Hailuo 02 carries the HOT label on VdoBloom, meaning its filtering is more permissive and flexible than most models in the catalog. Creators working on intense, dark, or edgy (though still lawful) concepts will hit fewer refusals here than on CENSORED-labeled models like Kling or Hailuo 2.3.

Does Hailuo 02 do text-to-video, image-to-video, or both?

Both in a single model. Write a prompt alone and it generates the scene from text, or upload a starting image and it animates from that frame. Either way you choose a 6 or 10 second duration at 768P or 1080P.

What is Hailuo 02 particularly good at?

Physics realism and camera control are its calling cards. Shots involving gravity, momentum, collisions, fluids, and deliberate camera movement tend to look more coherent than on models tuned mainly for static-scene aesthetics.

What resolutions does Hailuo 02 output?

Two tiers: 768P and 1080P. The 768P tier is the economical option for testing prompts and motion, while 1080P is the delivery-quality choice. Credit cost for each combination is shown before you generate.

Related models