Post-trained by fal · faster than real time
MiniMax H3 Max AI Video Generator
Turn a prompt or a photo into a 5 to 15 second video with synchronized sound. H3 Max renders a 5-second 768p clip in about two and a half seconds, so you can try an idea, change one line and try again without waiting.
- Text & image to video
- Native audio & lip sync
- 480p · 768p · 1080p
- 5–15 second clips
Create with MiniMax H3 Max
Write a prompt or upload a start frame, pick the length, resolution and shape, and see the exact credit price before you generate. Every clip comes back with its own soundtrack.
Aspect ratio
Length
Resolution
New accounts get free credits for a 5-second 768p video.
Made with MiniMax H3 Max
Your video will appear here.
MiniMax H3 Max video examples
Real H3 Max clips across the jobs people use it for most, each with the prompt that made it. Pick a category, then press "Use this prompt" to load it into the generator.
Product videos and ads
One product and one motion: a pour, a squeeze, a reveal. H3 Max keeps packaging and materials steady, and the foley comes with the shot, so a product clip is ready to cut into an ad.
Espresso macro
Text to videoMacro 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.
Burger from a product photo
Image to videoA gloved hand comes in from the right, presses the brioche crown down onto the stack and lifts it away again, dragging a long thread of melted cheese up with it before the thread snaps back onto the patty. Hold the slate board, the hard side light and the framing exactly as they are in the still, and keep the camera locked off throughout. Sound: the soft compression of the bun, cheese pulling, a faint sizzle off the patty, no music.
Sneaker sports ad
Text to videoA basketball player finishes lacing the red high-tops courtside, stands, and takes off up the court, the camera tracking low alongside him until he goes into a crossover at the free-throw line. Outdoor court at dusk under floodlights, chain-link fence behind, warm haze in the air. Sound: the ball on cracked asphalt, sneaker squeak and grit underfoot, the fence rattling as he passes, a couple of voices off court, no music.
3D cookie commercial
Text to video3D cookie commercial, cute cookie character, colorful background, playful typography, product packaging reveal, chocolate cookie close-up, vibrant studio lighting, fun advertising style.
Product explainer with cutaway
Text to videoCreate a polished 10-second photorealistic product explainer for a museum engineering supplier. The product is an unbranded seismic isolation plinth: a low, precision-machined circular pedestal with a pale limestone top, dark graphite base, and a thin vermilion accent ring. A delicate ancient celadon vase stands centered on top. The demonstration takes place in an immaculate museum conservation studio with cool gray walls, archival shelving softly out of focus, and warm directional daylight. 0–2 seconds: begin with a low three-quarter hero view and a slow, steady dolly toward the vase and plinth. The pedestal sits on a compact black earthquake simulation table. Crisp reflections and shallow depth of field emphasize the engineered materials. 2–6 seconds: the simulation table suddenly oscillates left and right with forceful, clearly visible earthquake motion. The black table and lower section of the plinth shift rapidly while the isolated limestone top glides in controlled counter-motion. The tall vase remains upright, calm, and nearly motionless, making the protection immediately understandable. A small shallow dish of water beside the vase stays level with only a faint ripple. Motion must feel mechanically credible, not magical. 6–8 seconds: the shaking decelerates and stops. The plinth smoothly recenters itself while the vase remains perfectly safe. The camera begins a subtle clockwise arc around the product. 8–10 seconds: during the arc, the dark outer casing becomes a refined semi-transparent technical cutaway, revealing crossed steel guide rails, circular bearings, and compact spring dampers returning to center beneath the stable top plate. End on a clean three-quarter close-up of the exposed mechanism and protected vase. Elegant museum-grade industrial design, celadon green, limestone cream, graphite, and vermilion palette; precise premium engineering-film aesthetic. Natural synchronized sound: low shake-table rumble, controlled mechanical sliding, then a soft settling click. No people, captions, logos, interface graphics, diagrams, or floating text.
All clips were generated with MiniMax H3 Max and are shown with their original prompts.
What is MiniMax H3 Max?
MiniMax H3 Max is a post-trained version of MiniMax H3, the open-weights video model MiniMax released in August 2026. fal's research team added a large amount of new training data and reinforcement learning aimed at two things: following the prompt more closely and looking better. H3 Max was announced on August 26, 2026.
The engine was built alongside the model, so speed is part of the design rather than a cut-down copy. fal's benchmark puts a 5-second 768p clip at about 2.5 seconds of render time, against roughly 53 seconds for standard H3 on the same job. In practice that means you can iterate on a shot the way you would tweak a photo.
Everything that made H3 useful carries over: synchronized audio in every clip, lip-synced dialogue, multiple shots inside one generation, and start and end frame control for image to video. On human-preference leaderboards H3 Max placed first for image to video at launch.
Tip: for the best mix of speed and quality, stay at 768p and generate 5 to 10 seconds — those are the defaults here. Move to 1080p for final renders.
- to render a 5 s 768p clip
- ≈2.5 s
- Design Arena image to video (Elo 1,341)
- #1
- per clip, up to 1080p
- 5–15 s
- dialogue, foley and music in every clip
- Audio on
MiniMax H3 Max vs MiniMax H3
H3 Max starts from H3, so the two share the same strengths. The differences are in speed, prompt adherence and where you can run them.
| What it is | MiniMax H3 Max | MiniMax H3 |
|---|---|---|
| What it is | Post-trained H3 by fal Research | MiniMax's open-weights base model |
| 5 s 768p render time | About 2.5 seconds | About 53 seconds |
| Image-to-video ranking (Design Arena) | #1 · Elo 1,341 | Elo 1,333 |
| Resolution and length | 480p, 768p, 1080p · 5–15 s | 768p, 2K · 4–15 s |
| Native audio and lip sync | Yes | Yes |
What you can do with H3 Max
Everything runs in the browser. No GPU, no install, and the price of each clip is shown before you press generate.
Text to video
Describe the subject, action, setting, camera and sound in plain words. Long, specific prompts work well — H3 Max follows beats in order and renders on-screen text legibly.
Image to video with start and end frames
Upload a product photo, portrait or illustration as the first frame. Add an optional end frame to control where the shot lands, which is handy for loops and reveals.
Native synchronized audio
Every clip comes with room tone, foley and music generated alongside the picture. Write dialogue in quotes and it is spoken with matching lip movement.
Six aspect ratios
21:9, 16:9, 4:3, 1:1, 3:4 and 9:16 for text to video, from cinema-wide to vertical. Image to video keeps the shape of your photo.
5 to 15 seconds, with cuts
Pick any length from 5 to 15 seconds. Longer clips can hold several shots of the same character if you describe them the same way in each beat.
480p, 768p or 1080p
Draft at 480p for the lowest cost, work at 768p — the resolution H3 Max is tuned for — and finish at 1080p. MP4 downloads have no watermark.
How to use MiniMax H3 Max
- 01
Choose text or image
Stay on Text to video and write your prompt, or switch to Image to video and upload a start frame (and an end frame if you want one).
- 02
Set length, resolution and shape
Pick 5 to 15 seconds, 480p to 1080p and an aspect ratio. The credit cost updates as you change settings, so there are no surprises.
- 03
Generate, watch and download
Press generate and the clip plays as soon as it is ready. Download the MP4, or tweak one line of the prompt and run it again.
MiniMax H3 Max pricing
You pay per second of video. A 5-second 768p clip costs 65 credits; 480p costs less and 1080p more. Failed generations are refunded automatically.
Basic
For trying ideas and the odd social post.
- 800 credits every month
- About 12 five-second 768p videos
- Text to video and image to video
- 480p to 1080p, 5 to 15 seconds
- Native sound, no watermark, commercial use
Pro
For creators who post video every week.
- 1,800 credits every month
- About 27 five-second 768p videos
- Text to video and image to video
- 480p to 1080p, 5 to 15 seconds
- Native sound, no watermark, commercial use
- Priority email support
Max
For marketers and teams making video at volume.
- 4,000 credits every month
- About 61 five-second 768p videos
- Text to video and image to video
- 480p to 1080p, 5 to 15 seconds
- Native sound, no watermark, commercial use
- Priority email support
FAQ
MiniMax H3 Max FAQ
Quick answers about the model, credits and what you can make. Still stuck? Email support@h3max-ai.com.
What is MiniMax H3 Max?
MiniMax H3 Max is a post-trained version of MiniMax's open-weights H3 video model, developed by fal Research and released in August 2026. It keeps H3's native audio and multi-shot abilities while following prompts more closely and rendering much faster.
How fast is H3 Max?
fal's benchmark measures about 2.5 seconds of render time for a 5-second 768p clip and about 15 seconds for a 15-second 768p clip. With upload and saving, most short clips on this page are ready within seconds; 1080p and 15-second clips take a little longer.
How much does a video cost?
Pricing is per second of output: 8 credits per second at 480p, 13 at 768p and 26 at 1080p. A 5-second 768p clip is 65 credits. The exact cost is shown next to the Generate button, and credits for a failed generation go straight back to your balance.
Can I turn a photo into a video?
Yes. Switch to Image to video, upload a JPG, PNG or WebP as the start frame and describe what should happen. You can also add an end frame so the shot finishes exactly where you want. The video keeps the aspect ratio of your photo.
Does H3 Max generate sound and speech?
Every clip includes audio generated together with the picture: ambience, sound effects and music. Put a line of dialogue in double quotes and it is spoken with lip sync. Describe the sound you want in the prompt, or write "no music" to leave it out.
Which resolutions, lengths and aspect ratios are supported?
480p, 768p and 1080p; any whole number of seconds from 5 to 15; and 21:9, 16:9, 4:3, 1:1, 3:4 or 9:16 for text to video. 768p is the resolution the model is tuned around, and 1080p is refined from it.
Can I use the videos commercially?
Yes. Videos you generate can be used in personal and commercial projects, such as ads, social posts and client work, as long as your prompts and uploaded images don't infringe anyone else's rights. Downloads have no watermark.
How do I write a good H3 Max prompt?
Cover five things: the subject, the action, the camera (lens and movement), the setting and light, and the sound. For longer clips, break the action into timed beats such as "0 to 3 s: …, 3 to 6 s: …". Use the examples above as starting points.
Is there a free trial?
Yes. Sign in with Google and your account gets enough free credits for a 5-second 768p video, so you can see the quality before you buy a plan or a credit pack.
Make your first MiniMax H3 Max video
Sign in with Google, get free credits and watch your first clip render in seconds.