Last verified: August 30, 2026.
Hailuo is MiniMax’s video generation family. It turns a text prompt or a still image into a short video clip, and it is available both as a consumer product and through the MiniMax API. Hailuo and MiniMax are not two companies: Hailuo is the product name, MiniMax is the company that builds it, which is why you will see both names attached to the same model.
This page is the starting point for everything Hailuo on this site: which model to use, what it costs, whether any of it is free, how to write prompts that work, and where Hailuo ends and the rest of the MiniMax range begins.
Independent-site notice: MiniMax-AI.chat is an independent educational website. It is not MiniMax’s official website and is not endorsed by or affiliated with MiniMax. Product names and trademarks belong to their respective owners. API access, interface availability, billing, and provider terms are controlled by MiniMax.
Start here
| If you want to… | Go to |
|---|---|
| Know whether you can use Hailuo without paying | Is Hailuo AI free? |
| Pick the right model for a new project | MiniMax-H3 — the current generation |
| Generate video from a still image | Hailuo 2.3 Fast |
| Control the camera in a shot | Hailuo video prompt guide |
| Keep one character consistent across shots | Hailuo S2V-01 |
| See every price in one table | MiniMax pricing |
The Hailuo models, and which one to use
MiniMax has moved Hailuo through two API generations. The distinction is not cosmetic: the two generations use different endpoints, different request payloads and different pricing, and a subscription that covers one does not necessarily cover the other.
| Model | Generation | Best for | Status |
|---|---|---|---|
| MiniMax-H3 | Video Generation V2 | New projects, 2K output, native stereo audio | Current |
| Hailuo 2.3 | V1 | Text-to-video and image-to-video at 768P or 1080P | Legacy |
| Hailuo 2.3 Fast | V1 | Faster image-to-video at lower cost | Legacy |
| Hailuo 02 | V1 | Earlier generation, still documented | Legacy |
| Hailuo S2V-01 | V1 | Character consistency from one reference image | Legacy |
| Hailuo Director | V1 | Explicit camera movement control | Legacy |
If you are starting today, price and prototype on H3 first. Do not carry a V1 payload, polling loop or pricing assumption across to it. The legacy pages remain on this site because a large number of existing integrations still run on V1 and need the reference.
What Hailuo costs
Hailuo is billed per generated clip, not per minute of compute. These are the published pay-as-you-go rates for Hailuo 2.3, verified August 13, 2026.
| Resolution | Duration | Price per clip |
|---|---|---|
| 768P | 6 seconds | $0.28 |
| 768P | 10 seconds | $0.56 |
| 1080P | 6 seconds | $0.49 |
Subscriptions handle video as a small daily allowance rather than out of the token pool: the Plus tier includes no Hailuo clips, Max includes 3 per day and Ultra includes 5 per day, and unused daily quota does not carry over. Because the highest tier works out to roughly $0.88 per clip against $0.28 pay as you go, a subscription is only worth it if you are also using its text token allowance. The reasoning is worked through in Is Hailuo AI free? and the tier detail is in the Token Plan guide.
Hailuo image to video
Image-to-video is the workflow most people actually want: you supply a still frame and Hailuo animates it, which gives you far more control over composition than describing a scene in words. Both Hailuo 2.3 and Hailuo 2.3 Fast accept an input image, and the Fast variant exists specifically for this path when turnaround matters more than maximum fidelity.
Two things decide whether the result is usable. The input image has to meet the documented resolution and aspect ratio requirements, or the request fails before generation. And the prompt has to describe motion, not content the image already shows — describing what is in the frame wastes the prompt, because the model can already see it.
Hailuo audio and voice
Hailuo itself generates video. Audio is a separate part of the MiniMax range, with one exception: MiniMax-H3 generates native stereo audio alongside the video, which the V1 Hailuo models do not.
For standalone speech, voice cloning and text-to-speech, see MiniMax Audio. If you are building a pipeline that needs both, generate video on H3 and narration through the speech API rather than expecting one call to do both.
The Hailuo app
MiniMax distributes a mobile application and a set of web products. Verified official download routes for every platform, plus which features are application-only versus available through the web or the API, are listed on our MiniMax app page.
One caution worth stating plainly: modified application packages advertising unlocked premium features or unlimited Hailuo credits are not MiniMax products, and in practice carry credential-stealing malware. No legitimate MiniMax plan grants unlimited video.
Writing prompts that work
Hailuo responds to explicit camera direction in a way that generic video prompts do not exploit. Commands such as push in, pull out, pan left, tracking shot and orbit are interpreted as instructions rather than description, and using them is the single change that most reduces wasted generations.
- Hailuo video prompt guide — camera commands and test-ready examples.
- Hailuo Director models — the models built specifically around camera movement control.
- Hailuo S2V-01 — keeping one character consistent across separate shots.
How Hailuo fits into MiniMax
Hailuo is one of several product lines under MiniMax. If you arrived here searching for Hailuo and want the wider picture, the starting points are the full model list, the API guide, and who owns MiniMax.
Prices, model status and quotas on this page reflect MiniMax documentation as verified on the date at the top. For account, billing and commercial decisions, use MiniMax’s official channels.