MiniMax H3 Max adds first-and-last-frame control to Hailuo 3's reference workflow — pin the start and end of a shot, or drive it with up to 9 images, 3 videos and 3 audio clips.
MiniMax H3 Max is the "Max" tier of the Hailuo 3 family, integrated here on 28 September 2026 alongside its Turbo sibling. Three endpoints sit behind one picker — text-to-video, image-to-video and reference-to-video — sharing the same panel and per-second pricing. What it adds over MiniMax H3 is a last-frame input for image-to-video and a cheaper 480p tier; what it gives up is the 2K tier, and the shortest clip is 5 seconds rather than 4.
The parts that cost you something: reference video is billed per input second at its own rate (40 credits at 480p, 100 at 768p), reference images beyond the second add 50 each, and reference audio is free. Audio has no off switch, submitted tasks cannot be cancelled, result links from the provider expire after 24 hours (we copy finished clips to our own storage), and we have not timed this model yet.
MiniMax's H3 Max generates 5–15 second clips with native audio at 480p or 768p, driven by text, a first frame (with an optional last frame), or a mixed set of images, video and audio references — no API key, no separate account.
Image-to-video on H3 Max accepts a first frame and an optional last frame. Hand it the opening and closing stills of a move — a product turning, a door opening, a face turning to camera — and describe the motion in between. With only a first frame it behaves like H3: the clip grows out of that image and keeps its aspect ratio.
Reference-to-video takes up to 9 images, 3 video clips and 3 audio tracks in a single request, referenced in the prompt as "Image 1", "Video 1", "Audio 1". Keep a face from a photo, a location from another, and drive dialogue from a voice track. The first two reference images are free; reference video adds billable input seconds; reference audio is free.
Two output tiers with per-second pricing shown before you generate: 35 credits per second at 480p, 55 at 768p. A 5-second 480p draft is 175 credits; rerun the same brief at 768p for delivery. Six aspect ratios from 9:16 vertical to 21:9 ultrawide in text and reference modes.
H3 Max runs from 5 to 15 seconds at a flat per-second rate, so a 12-second take costs exactly 12 seconds. Prompts up to 7,000 characters are kept with every generation, so a shot that works is reproducible.
Generated here on 28 September 2026, first take, exact prompts and settings shown.
A lighthouse keeper climbing a spiral iron staircase at dusk, lantern swinging, warm light sweeping the stone walls, wind and gulls outside, slow tilt up.
A jazz drummer in a small club, brushes on the snare, cymbals shimmering under a red spotlight, handheld camera slowly circling, room-tone applause.
It is the Hailuo 3 tier for shots with a known beginning and end, and for people who want a cheaper draft tier than H3 offers.
A packshot that ends on the label, a UI that ends on the final screen, a face that ends looking at camera — give it both stills and describe the move between them.
Nine images, three video clips and three audio tracks in one request, cited by order in the prompt — the same reference workflow as MiniMax H3.
35 credits per second is the cheapest way to iterate on a Hailuo 3 brief with references. Run the motion until it works, then rerun at 768p.
The ladder stops at 768p — pick MiniMax H3 for 2K. The shortest clip is 5 seconds, and audio has no off switch.
Same workspace, same credit pool — switching between any of these is one click.
| MiniMax H3 Max | MiniMax H3 | MiniMax H3 Max Turbo | |
|---|---|---|---|
| Resolutions | 480p · 768p | 768p · 2K | 480p · 768p |
| Clip length | 5–15s, any whole second | 4–15s, any whole second | 5–15s, any whole second |
| Image-to-video inputs | First frame + optional last frame | First frame only | First frame + optional last frame |
| Reference assets per request | Up to 9 images, 3 videos and 3 audio tracks | Up to 9 images, 3 videos and 3 audio tracks | Not offered |
| Credits per second, 480p | 35 | Not offered | 20 |
| Credits per second, 768p | 55 | 55 | 30 |
| Credits per second, 2K | Not offered | 90 | Not offered |
| Reference video, per input second | 40 at 480p · 100 at 768p | 55 at 768p · 90 at 2K | Not offered |
| Free reference images | First 2, then 50 each | First 5, then 30 each | Not offered |
| Credits for one clip, floor to ceiling | 175–825 for output alone; 2,675 fully loaded | 220–1,350 for output alone; 2,820 fully loaded | 100–450 |
| Sound | Native and always on | Native and always on | Native and always on |
Every figure in this table is read from our own model registry on 28 September 2026 — the same per-second rate tables that price your generations. Rates are per second of output; reference-video input seconds are added at the rate shown. A tier a model does not sell is marked "Not offered".
Compare MiniMax H3 Max with other video models — same workspace, one click to switch.
| Model | minimax-h3-max (MiniMax H3 Max, the Max tier of Hailuo 3) — integrated here on 28 September 2026 |
|---|---|
| Modes | Text-to-video, image-to-video from a first frame with an optional last frame, and reference-to-video from a mixed asset set — three endpoints behind one picker |
| Resolution | 480p and 768p — there is no 2K tier on Max (MiniMax H3 has it) |
| Duration | 5 to 15 seconds, any whole second (5 by default) |
| Aspect ratios | 16:9, 9:16, 1:1, 4:3, 3:4 and 21:9 in text and reference modes; image-to-video follows the first frame |
| Reference assets | Up to 9 images, 3 video clips and 3 audio tracks, 12 in total. Reference video runs at least 2 seconds per clip and 15 seconds combined. Cited in the prompt as "Image 1", "Video 1", "Audio 1" |
| Prompt length | Up to 7,000 characters |
| Audio | Native and always on — there is no audio toggle |
| Credits | Per second: 35 at 480p, 55 at 768p, the same rate in all three modes. Reference-video input seconds add 40 per second at 480p or 100 at 768p; reference images past the second add 50 each; reference audio is free |
| Credit range per clip | 175 for a 5-second 480p clip, 825 for a 15-second 768p clip. Loaded up with 15 seconds of reference video and 9 reference images, a 768p request tops out at 2,675 |
| Cancellation | Submitted tasks cannot be cancelled |
| Measured speed | Not measured — we have not run a timed benchmark on this model |
Three steps from a blank prompt box to a finished clip.
Open the model picker and choose MiniMax H3 Max. The panel shows its credit cost, resolution options, and duration limits before you spend anything.
Write the subject, the action, and how the camera behaves. Attach a first frame if you want the clip to start from a specific image.
Download the clip, extend it, or send it through further editing. Prompt, model, and settings stay attached in your history.
One account, the full image and video generation workflow.
Run MiniMax H3 Max on the free tier's daily credits, or upgrade for premium models, faster generation, and 4K output.
Perfect for getting started
Billed annually · $240/year
Billed annually · $468/year
Shots where you know how they should start and end — first-and-last-frame image-to-video — plus the multimodal reference workflow Hailuo 3 is known for. Output is 5–15 seconds at 480p or 768p with native audio.
35 credits per output second at 480p and 55 at 768p, so a 5-second 480p clip is 175 credits and a 5-second 768p clip is 275. Reference video adds 40 credits per input second at 480p or 100 at 768p; the first two reference images are free and each further image adds 50; reference audio is free. The exact total is shown before you generate.
H3 Max adds a last-frame input and a cheaper 480p tier, but tops out at 768p — only MiniMax H3 offers 2K. H3 Max Turbo is the same 5–15 second engine at about half the price, without the reference-asset mode.
Up to 9 images, 3 video clips (at least 2 seconds each, 15 seconds combined) and 3 audio tracks, at most 12 assets in total. Refer to them in the prompt by order: "Image 1", "Video 1", "Audio 1".
Whole seconds from 5 to 15, at 480p or 768p. Text and reference modes offer 16:9, 21:9, 4:3, 1:1, 3:4 and 9:16; image-to-video follows the ratio of your first frame.
No. H3 Max generates native audio on every run and there is no toggle, so the rate you see is the only rate. Mute it in your edit if you need a silent plate.
One-click tools for everything around generation — polish, restyle, and share.
