Hailuo 2.3 is the latest video model from MiniMax, and it doubles down on the thing the Hailuo line is best known for: convincing, physics-aware character and object motion. Version 2.3 is dedicated entirely to image-to-video — you supply the picture, it supplies the movement — which lets the model pour all of its capacity into animation quality rather than splitting attention with text-only generation.
Flux 3 AI hosts two Hailuo 2.3 tiers. Pro is the fidelity pick for finished deliverables, while Standard renders at a gentler credit price for everyday clips and experiments. Both start from your uploaded image, take a reference image to reinforce the subject, and expose style presets so a photo can move like film, anime, or something in between.
Every clip is 6 or 10 seconds in 16:9 widescreen or 9:16 vertical, rendered up to 1080p with a lighter 768p option for drafts. If your source is a strong still — a portrait, a product shot, an illustration — Hailuo 2.3 is arguably the sharpest tool on the platform for bringing it to life; when you need to generate from words alone, its sibling Hailuo 02 covers text-to-video.
See what's possible on Flux 3 AI
Sample generations from the platform — write your own prompt to see what Hailuo 2.3 does with it.
Hailuo 2.3 vs similar AI video models
Specs from the live catalog — every model below is available on the same account.
| Spec | Hailuo 2.3 | Hailuo 02 | Kling 2.5 Turbo |
|---|---|---|---|
| Max clip length | 10s | 10s | 10s |
| Durations | 6s, 10s | 6s, 10s | 5s, 10s |
| Resolutions | 768p · 1080p | 720p · 768p · 1080p | 720p · 1080p |
| Aspect ratios | 2 | 2 | 3 |
| Audio | |||
| Image to video | |||
| Reference images | up to 1 | up to 1 | up to 1 |
| Keyframes | |||
| Negative prompt |
Capabilities reflect what each model supports on Flux 3 AI today.
What makes Hailuo 2.3 stand out
Motion-first design
A dedicated image-to-video architecture concentrates on natural body mechanics, cloth, and camera feel.
Pro and Standard pricing
Standard keeps experimentation affordable; Pro steps up detail and stability for client-ready output.
Reference reinforcement
Add a reference image alongside your source frame so faces and products hold their identity in motion.
Full HD delivery
Render finals at 1080p, or use the 768p option to preview movement before spending on the big pass.
Style presets
Built-in looks shift the same still between cinematic, animated, and stylized motion treatments.
Feed-ready framing
Native 16:9 and 9:16 outputs slot straight into YouTube, Reels, TikTok, and Shorts without re-cropping.
Popular use cases
Why creators run Hailuo 2.3 on Flux 3 AI
- Feed the same still to Hailuo 2.3 and Kling, then pick the better animation side by side.
- Watermark-free downloads make every Hailuo clip ready for client delivery straight away.
- Sign-up credits are enough to test both Pro and Standard tiers before choosing a plan.
Hailuo 2.3 specs & capabilities
Output
- Clip duration6, 10 seconds
- Resolutions768p · 1080p
- Aspect ratios16:9 · 9:16
- Generation modesImage to Video
Controls
- Image to video (start frame)
- Reference images
- First & last frame keyframes
- Audio generation
- Negative prompt
- Video input / editing
- Extend video
Which Hailuo 2.3 version should you use?
Every version below is selectable in the Create studio on the same credits.
Hailuo 2.3 Pro (I2V)
The fidelity tier for finished work. It animates your uploaded image into a 6- or 10-second clip at 768p or 1080p in 16:9 or 9:16, with one reference image and style presets to hold the subject and look. Use it for client deliverables and anything you plan to publish.
Hailuo 2.3 Standard (I2V)
The economical sibling with identical mechanics — same 6- and 10-second durations, 768p and 1080p outputs, both aspect ratios, reference support, and presets — at a lower credit cost per render. Start motion tests and everyday clips here, then rerun the keeper on Pro if it needs extra detail.
Hailuo 2.3 prompt examples
Copy one as a starting point, or send it straight to the Create studio.
“Animate this portrait: she turns her head toward the camera, hair lifting in a light breeze, a slow smile forming, subtle handheld drift, soft window light.”
Tests lifelike facial motion and hair physics.
“Bring this watch photo to life: the face tilts to catch the light, second hand ticking, camera pushing in slowly, reflections sliding across the sapphire glass.”
Micro-motion and reflection handling on a static product shot.
“From this illustration, the dragon unfurls its wings and exhales drifting embers, smoke curling upward, gentle parallax between the foreground and the painted sky.”
Shows stylized animation of 2D artwork with parallax depth.
Create with Hailuo 2.3 in three steps
Describe your idea
Write a prompt or upload a starting image — the more specific the scene, motion, and style, the better Hailuo 2.3 performs.
Pick Hailuo 2.3 & settings
Choose duration, resolution, and aspect ratio. The studio shows the exact credit cost before you generate.
Generate & download
Hailuo 2.3 renders in the cloud — track progress in your library, then download watermark-free or share with a link.
Hailuo 2.3 questions, answered
What is Hailuo 2.3?
Hailuo 2.3 is MiniMax’s newest video model, specialized in image-to-video generation. Upload a still image on Flux 3 AI, describe the motion you want, and it returns a 6- or 10-second animated clip at up to 1080p in Pro or Standard quality.
Can Hailuo 2.3 generate video from text only?
No — version 2.3 requires a starting image and focuses exclusively on animating it. For prompt-only generation from the same family, use Hailuo 02, which offers both text-to-video and image-to-video and is also available on Flux 3 AI.
Hailuo 2.3 Pro vs Standard — what changes?
Pro produces the higher-fidelity result and is the tier to use for final deliverables; Standard costs fewer credits and renders everyday content or motion tests. Durations, aspect ratios, and reference support are identical between the two.
What image should I upload for the best Hailuo 2.3 results?
Sharp, well-lit images with a clear subject animate best. Because output tops out at 1080p, sources at or above that resolution preserve detail, and adding a reference image of the same subject helps the model keep identity consistent through movement.
How long and what size are Hailuo 2.3 videos?
Clips run 6 or 10 seconds, in 16:9 or 9:16, at 768p or 1080p. That covers horizontal platforms and vertical feeds directly from generation, with the smaller resolution serving as a quick, cheaper preview step.
Why choose Hailuo 2.3 over other image-to-video models?
The Hailuo line has a reputation for lifelike human movement and believable physics, and 2.3 is its most refined expression of that strength. On Flux 3 AI you can run the same still through Hailuo 2.3 and a competitor like Kling or Seedance to compare directly.
Does Hailuo 2.3 cost money to try?
New Flux 3 AI accounts include free sign-up credits usable on Hailuo 2.3 and the rest of the catalog. Ongoing use runs on plans or credit packs from the pricing page, and generated videos always download watermark-free.
Do I need a MiniMax account to use Hailuo 2.3?
No — Hailuo 2.3 runs entirely through your Flux 3 AI account. There is no separate MiniMax registration, subscription, or key involved; the same login and credit balance you use for other models on the platform covers both Hailuo 2.3 tiers.
Can I access Hailuo 2.3 through an API?
Yes. Beyond the browser studio, Flux 3 AI provides API access for developers, which lets you build Hailuo 2.3 image-to-video generation into your own tools and automations. See the API section on the homepage for how to get started.