Hailuo in Doppelshot
Hailuo is the video model family of MiniMax. Doppelshot runs it everywhere a clip must be cheap and steady: the one-click effects at 5 credits, the 360 orbit, Make it longer, and the H3 Max reference engine behind Recast, Reframe and Put me in this scene.
What it is
- Reference-to-video with several people: everyone in the photo moves, the same place throughout.
- H3 Max takes a video as reference and replaces, reframes or inserts a person.
- The cheapest route to a clip that keeps the face.
What Doppelshot makes with it
- Levitate, set on fire, money rain, bullet time and the other effects: 5 credits.
- Recast (you in any clip), Reframe my video, Put me in this scene, on H3 Max.
- Make it longer: 10 credits per added 10 seconds.
Engines and prices
| Engine | Kind | Price | Cards |
|---|---|---|---|
| Hailuo | Scene video | 5 credits | Couple video, Bring it to life, Levitate +19 |
| Insert (H3 Max) | Motion video | 5 credits | Put me in this video, Reframe my video, Put me in this scene +1 |
| Orbit (H3 360 LoRA) | Scene video | 5 credits | Orbit |
| Hailuo extendRecommended | Make it longer | 10 credits | Make it longer |
| Reframe (H3 Max) | Motion video | 12 credits | Reframe my video, Put me in this video, Put me in this scene +1 |
| Recast (H3 Max) | Motion video | 35 credits | Put me in this video, Reframe my video, Put me in this scene +1 |
Checked against the catalog on 2026-10-21.
Cards that run on it
Open a card, drop your selfie, and pick this engine in the Studio when it is not the default.
Video5 credits
Video5 credits
Video5 credits
Video5 credits
Video5 credits
Video5 credits
Video5 credits
Video5 creditsQuestions
Why are the effects so cheap?
They are short 5-second clips on Hailuo, enough for a social post. The scene cards on Veo, Kling or Seedance cost more and do more.
What does H3 Max do?
It reads a reference video and puts you in it: your face on the actor, a new framing, a scene inserted at a timestamp.
Try it on your selfieThe first preview is free, no card needed.
Other models
from 7 creditsSeedance 2.5by ByteDanceByteDance's video model: multi-shot scenes, sound, a steady face.from 11 creditsKlingby KuaishouKuaishou's video models: Kling 4.0 up to 30 s, Kling 3 Pro 5 to 15 s, O3 extends a clip.from 5 creditsVeo 3.1by GoogleGoogle's video model: cinematic light and motion, with sound. A Lite tier at 5 credits.from 10 creditsOmniHuman 2.0by ByteDanceByteDance's talking model: lip sync, gestures, a direction you give it.