Camera moves you can direct
MiniMax H3 Max: A low tracking shot stays locked on a scooter as it races downhill, carrying speed, lean, background motion, and engine sound through the turn.
MiniMax H3 Max AI Video Generator
Create 5–15 second videos from a text prompt or opening frame, with an optional ending frame to guide the finish. H3 Max renders 480P or 768P at 24 FPS with synchronized audio in the same generation.
480P / 768P · 5–15 seconds
Native synchronized audio
Text or first/last frames
0/20000
01
H3 Max · Introduction
Overview, evidence, benchmarks, context.
MiniMax H3 Max · 768P · Native Audio
Faster renders. Sharper direction.
H3 Max is a fal post-trained variant of the open-weight H3 base, built for fast text-to-video, first-to-last-frame animation, and structured long prompts with native synchronized audio. When testing H3 Max, record the inputs, duration, and review criteria together. H3Max is an independent tool, not an official MiniMax or fal website.
Text to VideoImage to Video768P5–15sNative Audio
≈2.5sbackend inference for a 5s 768P clip; queue and prompt expansion add to total wait time.
Demo source: fal MiniMax H3 Max · 2026-08-30
Model selection guide
MiniMax H3 Max for speed; standard H3 for broader inputs and editing.
The homepage workspace defaults to H3 Max: 480P/768P, 5–15 seconds, text or first/last-frame input, and native synchronized audio. Keep H3 Max selected for that focused path. The examples below belong to standard H3, which covers richer references, video editing, and output up to 2K.
02
H3 Max · How to use
Prompt, settings, review, iteration.
How to use
How to use MiniMax H3 Max in three controlled steps
Treat H3 Max as a short production loop: define the shot, choose output constraints, then review the H3 Max result with the same checklist. Use H3 Max with one deliberate change per iteration. A consistent H3 Max process makes prompt changes easier to compare.
- 01
Write the shot or add a first frame
Describe the subject, action, setting, camera, lighting, and sound. H3 Max works best when each visible beat is concrete. For image-to-video, upload an opening frame and add an ending frame only when the finish needs explicit direction.
- 02
Choose settings and estimate credits
Select 480P or 768P, a 5–15 second duration, and an aspect ratio. H3 Max costs 10 credits per second, so every H3 Max test has a visible cost before submission.
- 03
Play, inspect, and iterate
Unmute the H3 Max result and inspect action order, subject identity, camera motion, and synchronization. Change one variable at a time before running H3 Max again.
Switch to standard H3 for multi-reference input
H3 Max focuses on fast text or first/last-frame generation. Standard H3 is the better fit when images, video clips, and audio must be combined as references in one request. This boundary keeps standard H3 features from being attributed to H3 Max. These examples preserve that multi-reference workflow so the two model variants are not confused.
Input
Image 1
Image 2
Image 3
Video 1
Audio 1
Output
Prompt:Using the character actions and camera language from @Video 1, generate a combat sequence featuring the characters from @Image 1 and @Image 2, set against the background of @Image 3. The fight choreography and visual style should emulate the pixel-art, side-scrolling combat of the classic game Contra, including retro movement rhythms and arcade-style impact beats. Apply the background music from @Audio 1, and synchronize additional combat sound effects—hits, jumps, and weapon actions—with the on-screen fighting motions to enhance intensity and immersion.
Use standard H3 to revise an existing clip
H3 Max currently covers text-to-video and first/last-frame image-to-video. Replacing a character, object, background, or line of dialogue in an existing clip belongs to the standard H3 workflow. The example keeps that editing workflow distinct from H3 Max.
Input
Image 1
Video 1
Output
Prompt:Replace the outfit worn by the runway model in @Video 1 with the clothing shown in @Image 1, ensuring natural fit, consistent fabric behavior, and seamless integration with the model’s movement, lighting, and original camera motion.
03
H3 Max · Features
Capabilities, prompts, editing, models.
Capabilities
Six MiniMax H3 Max tasks that test different kinds of control
These MiniMax H3 Max capability cards isolate camera direction, frame control, identity, audio, style, and event order.
Two stills become one shot
MiniMax H3 Max: An opening frame and optional closing frame become one continuous journey, with motion and synchronized sound generated between them.
The same character, shot after shot
MiniMax H3 Max: One generation moves through six locations while preserving the character’s face, hair, clothing, and proportions as lighting changes.
Sound generated with the picture
MiniMax H3 Max: Room tone, foley, music, dialogue, and ambience arrive in the same pass, synchronized to what happens on screen.
One look held across every shot
MiniMax H3 Max: A six-shot animated Greek vase sequence holds the same palette, linework, gold leaf, and hand-set lettering throughout.
It hits the beats in order
MiniMax H3 Max: Detailed beats arrive in sequence, including requested transformations and legible on-screen words, without losing the visual direction.
Made with MiniMax H3 Max
Reusable MiniMax H3 Max prompts, from short direction to full shot briefs
Each card pairs the original English prompt with its resulting video. Every MiniMax H3 Max example opens independently, keeping folded prompt text out of the initial reading flow. Turn on sound to inspect native audio, copy and adapt the prompt, or send it back to the homepage workspace with Try.
Shot Brief
One continuous take, no cuts
MiniMax H3 Max result: A dense 15-second cinematic brief with timed beats, physical camera direction, color design, and native sound.
View full prompt
Pixel Art
Playable-looking game footage
MiniMax H3 Max result: A long structured prompt coordinates gameplay rules, animation states, parallax scenery, UI, music, and sound effects.
View full prompt
Dialogue
A line to camera, lip synced
MiniMax H3 Max result: A concise dialogue prompt combines performance, lens, lighting, voice, and kitchen ambience.
View full prompt
Stop Motion
Claymation frog with pancakes
MiniMax H3 Max result: Material texture, stepped movement, staging, music, and tiny foley cues stay coherent in one pass.
View full prompt
Slow Motion
Hummingbird at a trumpet flower
MiniMax H3 Max result: A compact nature prompt directs timing, light, optics, detail, and a layered garden soundscape.
View full prompt
Product
Espresso, macro
MiniMax H3 Max result: Product texture, macro optics, steam, reflections, and cafe sound are directed in a single paragraph.
View full prompt
Features
Everything an AI Video Generator Should Do
MiniMax H3 unifies generation, reference, and editing in one model. These complete features are available after purchasing a plan.
Use these features to decide whether MiniMax H3 Max fits the shot, then move to standard H3 only when broader inputs are required.
01
Text to Video
Describe a scene in plain words — up to 7,000 characters — and MiniMax H3 turns it into a 5–15 second video with motion, camera work, and sound. The more film language you use, the more directed the result feels.
02
Image to Video
Animate any photo, product shot, or illustration. Set a first and last frame to control exactly how a shot begins and ends, and H3 fills in smooth, natural motion between them at up to 1440p.
03
Native Audio Built In
Every video comes out with two-channel stereo sound — ambience, effects, music, and lip-synced dialogue. No silent clips, no separate sound design step, no extra tools.
04
Edit Video with a Sentence
Upload a clip and tell H3 what to change: swap a character, remove an object, replace the background, relight the scene, or rewrite the dialogue. Unedited content stays stable frame to frame.
05
Reference Anything
Combine up to 9 images, 3 video clips, and 3 audio tracks in one request. Keep a character's face consistent, copy choreography from a dance clip, or clone a voice from a sample.
06
Every Aspect Ratio
Generate in 21:9, 16:9, 4:3, 1:1, 3:4, or 9:16 at 24 FPS. One AI video generator covers cinema-wide trailers and vertical phone-first content alike.
04
H3 Max · User reviews
Verified comments from creators.
User reviews
Feedback from real creative workflows
These six original comments describe practical MiniMax H3 Max workflows. They are shown without star ratings, audience totals, or unverified aggregate claims.
M
Marcus Deleon
Indie Game Developer
I made a full character PV for my game with MiniMax H3 in one afternoon. The menu UI stayed readable in every frame and the character never went off-model. That used to cost me a contractor and three weeks.
A
Aiko Tanabe
Animation Studio Lead
We tested every AI video generator on the market for stylized work. H3 is the only one that held our anime style across cuts. We now use it for pitch reels and animatics on every project.
P
Priya Raghavan
E-commerce Brand Owner
I uploaded four product photos and got a listing video with music and a voiceover the same day. My click-through rate on the new listings is up 38%. H3 even got the label text on my packaging right.
D
Danielle Whitfield
Social Media Manager
I ship five vertical videos a week for three brands. MiniMax H3 gets me from brief to draft in under an hour, with sound already synced. My old workflow needed a videographer and two review rounds.
T
Tomás Herrera
Freelance Video Editor
The editing side is what sold me. A client wanted the background of an interview changed — I typed one sentence into MiniMax H3 and the lighting on the subject matched the new scene automatically.
O
Oliver Bennett
Film Student
I previsualized my entire short film in MiniMax H3 before we shot a single frame. Rack focus, match cuts, even the title cards — it understands director language, not just keywords.
05
FAQ
Inputs, outputs, credits, boundaries.
FAQ
MiniMax H3 Max FAQ
Practical answers about MiniMax H3 Max inputs, output settings, the speed benchmark, credits, prompt structure, and the boundary between MiniMax H3 Max and standard H3.
MiniMax H3 Max is designed for fast text-to-video and first/last-frame image-to-video at 480P or 768P for 5–15 seconds with native synchronized audio. Standard H3 is the choice for multi-reference input, existing-video editing, and output up to 2K.
H3Max support: hi@h3max.io
06
CTA
Turn evidence into generation.
Start creating
Create your first MiniMax H3 Max video
Describe the subject, action, camera, and sound on the homepage, then choose 480P or 768P and a 5–15 second duration. MiniMax H3 Max also accepts an opening and optional ending frame; required credits are shown before submission.
Choose a plan after signup
Native audio on every video
Up to 1440p cinematic output
