MiniMax H3 Max
Generate MiniMax H3 Max video from text or a start and optional end frame at 480p or 768p, with synchronized native audio and exact credit quotes.
Also consider: Use standard MiniMax H3 when you need 2K output or multimodal reference-to-video; choose H3 Max for faster 480p or 768p iteration.
Switch anytime · prompt preserved · one shared credit balance
Your video will appear here
Shot plan for this route
Choose H3 Max when turnaround speed, prompt adherence, and a finished 768p shot with sound matter more than 2K delivery or mixed image, video, and audio references. Use Text mode to invent the full scene, or Image mode to preserve a chosen opening composition and optionally define the ending frame.
- 01
Choose Text or Image
Start from a written brief, or switch to Image mode and upload the opening frame plus an optional final frame.
- 02
Write the shot and sound
Describe subject, visible action, one camera move, lighting, and the dialogue or ambience that belongs in the scene.
- 03
Pick duration and resolution
Choose 5, 10, or 15 seconds and compare the exact 480p or 768p credit quote before submitting.
- 04
Review the first result
Check prompt adherence, motion, continuity, and audio sync, then revise one variable at a time.
MiniMax H3 Max reel
Reference clips for action, camera, and sound continuity. Treat them as shot boards, not final deliverables.
Model brief
What MiniMax H3 Max is for
MiniMax H3 Max is fal Research's post-trained version of the open-weight MiniMax H3 base model. It is tuned for stronger prompt adherence and aesthetics, then co-optimized for fal's inference stack. This workspace connects the two live launch routes: text-to-video and image-to-video. Both generate synchronized audio; image mode can take a first frame and an optional last frame.
Production notes
Keep the comparison tied to controls you can actually set in the generator.
- Two live input modes
- Switch between text-to-video and first/last-frame image-to-video without leaving this page.
- 5 to 15 seconds
- The generator offers 5, 10, and 15-second clips, matching the verified endpoint range.
- 480p or 768p
- Use 480p for lower-cost drafts or 768p for the model's tuned output tier; standard H3 remains the 2K option.
- Native synchronized audio
- Dialogue, ambience, effects, and music are generated with the picture in the same pass, so audio is always on for this route.
- Balanced prompt expansion
- Requests use fal's balanced prompt-expansion mode to improve the brief without adding the longer quality-mode rewrite delay.
Workspace facts
Live constraints for MiniMax H3 Max on this route: endpoints, duration/resolution wiring, credit math, and provider pricing sources.
- Live endpoints on this page
- MiniMax H3 Max opens through 2 linked endpoints in the MiniMax H3 generator: minimax/h3-max/text-to-video; minimax/h3-max/image-to-video. Available modes here: text to video, image to video. Durations offered in the UI: 5, 10, 15 seconds. Resolutions wired for this model: 480p, 768p.
- Credit cost in this workspace
- A 5s 480p run with MiniMax H3 Max costs 65 credits here (provider estimate about $0.250 before margin rules). No reference-to-video sibling is linked for this keyword here. Supported resolutions for this model: 480p, 768p. Per-credit value uses the lowest paid MiniMax H3 plan (about $0.012).
- Provider pricing source
- Upstream video pricing is recorded per output second by resolution (480p: $0.0500/s; 768p: $0.0800/s). Checked 2026-08-28. Primary source: https://fal.ai/models/minimax/h3-max/text-to-video.
- Compared with the MiniMax H3 default
- Home default remains H3 Max Turbo · Text to Video. This page locks MiniMax H3 Max for a controlled comparison in the same prompt box and credit estimator.
- Honest limits
- Reference-to-video is not offered for this keyword here.
MiniMax H3 Max shot prompts
Write the shot in production order. Copy a brief, replace the subject, keep camera and sound concrete.
Dialogue to camera
Keep the line short, name the performance beat, then anchor camera, light, and room sound.
A chef in her forties looks straight into the camera in a bright open kitchen and says, ‘The secret is you never crowd the pan.’ She smiles and turns back to the stove. Clean daylight from a window, white tile, flour on her apron, a strand of hair loose. Medium close-up, 50mm, slight handheld. Sound: her voice close and clear, a gentle sizzle, a knife on a board off screen.
Stop-motion art direction
State the material, frame cadence, visible imperfections, action sequence, and matching sound cues.
Stop-motion claymation: a round green frog in a knitted scarf carries a stack of pancakes across a tiny kitchen set, sets it down, then looks at camera and blinks. Visible fingerprints in the clay, felt walls, warm practical lamps, stepped 12fps motion. Sound: playful pizzicato strings, soft clay squeaks, a plate clink.
Slow-motion nature detail
Describe subject motion, lens behavior, light, atmosphere, and a small set of realistic sounds.
Slow-motion close-up of a hummingbird hovering at a red trumpet flower in a sunlit garden, wings a blur, tongue flicking into the bloom. Backlit at golden hour with bokeh highlights, pollen and dust in the air, one bruised petal. Long lens, very shallow focus. Sound: rapid wing hum, garden birdsong, a distant sprinkler.
Macro product shot
A simple physical process, one lens choice, and close sound cues give a short product clip a clear arc.
Macro shot of espresso pulling from a chrome portafilter into a warm glass, thick golden crema blooming and swirling. Bright morning light through a cafe window, steam curling, water spots on the chrome, a few coffee grounds on the tray. 100mm macro, shallow focus, slight handheld drift. Sound: grinder winding down, the hiss of the pump, liquid pattering into glass, cafe chatter behind.
Primary sources and documentation
Endpoint capabilities, limits, and pricing on this page are tied to these upstream sources.
- falMiniMax H3 Max launch and benchmarks
fal's launch page documents the post-training, endpoint scope, native audio, speed claims, resolution range, duration range, and comparison with standard H3.
- fal API DocsH3 Max text-to-video API
Official schema for prompt, duration, 480P/768P resolution, aspect ratio, safety checking, and prompt expansion.
- fal API DocsH3 Max image-to-video API
Official schema for first-frame image input, optional last frame, duration, resolution, and output video.
MiniMax H3 Max FAQ
What is MiniMax H3 Max?+
MiniMax H3 Max is a fal Research post-trained variant of the open-weight MiniMax H3 base model. It is optimized for prompt adherence, aesthetics, and fast inference on fal's serving stack; it is not the same endpoint as standard MiniMax H3.
Which MiniMax H3 Max modes are live here?+
Text-to-video and image-to-video are live. Image mode accepts a required first frame and an optional last frame. A MiniMax H3 Max reference-to-video endpoint is not advertised on this page until it is publicly available and verified.
What resolutions and durations does H3 Max support?+
This generator offers 480p and 768p at 5, 10, or 15 seconds. Text mode supports the site's six aspect ratios; image mode follows the aspect ratio of the uploaded first frame.
Does MiniMax H3 Max generate audio?+
Yes. H3 Max returns synchronized native audio with every video, so this page does not show a misleading audio-off switch. Include dialogue, ambience, effects, or music direction in the prompt.
How is H3 Max different from standard MiniMax H3?+
H3 Max is the faster 480p/768p route tuned by fal. Standard MiniMax H3 remains the better fit when you need 2K output or mixed image, video, and audio references.
How many credits does H3 Max use?+
H3 Max uses 13 credits per output second at 480p and 20 credits per output second at 768p on this site. The generator recalculates the exact total before every request.