MiniMax H3 Max Turbo
The speed tier of the H3 video family — 5–15s clips with native stereo audio at up to 1080P, tuned for rapid iteration and high-volume generation.
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax-h3-max-turbo-text-to-video",
"prompt": "Aerial drone shot over a misty mountain valley at golden hour",
"duration": "5s",
"resolution": "480P",
"aspect_ratio": "16:9"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Overview
What is MiniMax H3 Max Turbo
MiniMax H3 Max Turbo is a speed-optimized video generation model that turns text prompts or a still image into 5–15 second clips with native synchronized audio, at 480P, 768P, or 1080P across six aspect ratios. It distills H3 Max quality into a faster, cheaper tier built for rapid iteration and high-volume production.
Running it privately on Venice
On Venice, MiniMax H3 Max Turbo runs under a zero-retention privacy tier: your prompts and clips are not stored, profiled, or used for training, and there is no personal generation history tied to your identity. The model is also uncensored, so creative prompts are not filtered the way they are on Big-Tech video tools. You pay per clip in credits from $0.11 — no subscription required — and both the text-to-video and image-to-video variants are available through the app or the Venice API.
Specifications
Datasheet
- Maker
- MiniMax
- Modality
- Text-to-video and image-to-video, both with native audio
- Open weights
- No — proprietary distilled checkpoint
- License
- Proprietary (distilled checkpoint; underlying H3 base reported open-weight)
- Clip lengths
- 5s – 15s
- Resolutions
- 480P, 768P, 1080P
- Mode
- text-to-video
- Audio
- Yes
- Prompt limit
- 10,000 chars
- Input images
- One still image for image-to-video; optional first-frame and last-frame guidance images for controlled transitions (accepted formats and size limits not published)
- Released
- September 2, 2026
- Aspect ratios
- 16:9, 21:9, 4:3, 1:1, 3:4, 9:16
- Privacy on Venice
- Private — zero retention
- Available on Venice since
- Sep 2026
Assessment
Strengths and limitations
- Built for speed: third-party testing in September 2026 ranked H3 Max Turbo the fastest of five speed-tier video models, with renders that finish before a 15-second clip finishes playing.
- Native stereo audio is generated alongside the picture — effects, room tone, and dialogue arrive in sync rather than being dubbed on afterward.
- Strong prompt adherence and visual character inherited from H3 Max, which tops third-party image-to-video leaderboards.
- Full creative envelope: 5–15 second clips, 480P to 1080P, and six aspect ratios from 21:9 cinematic to 9:16 vertical.
- First-and-last-frame control lets you lock the start and end of a shot and let the model invent the motion between.
- Runs uncensored on Venice with zero retention: your prompts are not stored or used for training.
- Closed weights: the Turbo checkpoint is proprietary and cannot be self-hosted or fine-tuned, even though the underlying H3 base model is reported to be open.
- Distillation costs some fine detail: side-by-side comparisons show Turbo softening micro-detail (a match flame, a watch bezel) that H3 Max resolves sharply.
- No end-to-end encryption or TEE on Venice: privacy comes from zero retention of prompts and outputs, not encrypted compute.
- Third-party benchmark aggregators score it below flagship models like Seedance 2.5 and Wan 3.0 on overall quality.
- Third-party docs recommend 5–10 seconds as the sweet spot, with longer clips more prone to drift.
Use cases
What it is good for
- 01Rapid creative iteration — block a shot, watch it, adjust the prompt, watch it again at 480P cost.
- 02High-volume UGC and ecommerce clips where per-second cost and turnaround dominate.
- 03Previz and animatics for film and advertising, with 21:9 framing and native audio.
- 04Vertical social content (9:16) with synchronized sound for Reels, Shorts, and TikTok.
- 05Controlled transitions: set a first and last frame and let the model generate the motion between them.
- 06Animating a product photo or character still into motion via the image-to-video variant.
Prompting
Getting better results
Front-load subject, camera move, and lighting in the first sentences — the 10,000-character limit is generous, but adherence is strongest on what comes early.
Iterate at 480P · 5s ($0.11 per clip), then re-render only the winning prompt at 1080P.
Describe the sound you want explicitly ("rain on pavement, distant thunder") — audio is generated with the video, so naming it changes it.
Use the image-to-video variant with a character or product still when identity must stay consistent across clips.
Supply first and last frames to lock a transition, then let the model invent the motion between the two images.
Pick 9:16 for Reels and TikTok, 21:9 for cinematic reels — aspect ratio is a generation parameter here, not a post-crop.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
Slow aerial drone shot gliding over a misty mountain valley at golden hour, sunlight piercing clouds onto a winding river, ancient pine forests on either side, ultra-smooth motion, professional color grading, atmospheric haze, 4K
Calm ocean waves rolling onto a black-sand beach at sunrise, golden light on wet sand, foam dissolving into the shore, a single silhouetted figure at the waterline, smooth continuous forward push, serene cinematic atmosphere
A slow tracking shot through a rain-soaked Tokyo street at night, neon reflecting in puddles, steam rising from a food cart, a person with a translucent umbrella, shallow depth of field, teal-and-orange grade, smooth steady camera
Alternatives
How it compares
| Model | Best for | Max duration | Max resolution | Native audio |
|---|---|---|---|---|
| MiniMax H3 Max Turbo | Fast drafts & high-volume clips | 15s | 1080P | Yes |
| MiniMax H3 Max | Maximum H3 quality | 15s | 1080P | Yes |
| Grok Imagine 1.5 | Fast social clips | 15s | 1080p | Yes |
| Wan 2.7 Enhanced | Open-weights workflow | 15s | 1080p | Yes |
| Vidu Q3 | Longest clips | 16s | 1080p | Yes |
Pick MiniMax H3 Max Turbo when throughput is the brief — rapid iteration, previz, and UGC volume — and you want H3-family prompt adherence with native audio at the lowest per-clip price on Venice. Choose H3 Max when a final render must resolve every detail.
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id.
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax-h3-max-turbo-text-to-video",
"prompt": "Aerial drone shot over a misty mountain valley at golden hour",
"duration": "5s",
"resolution": "480P",
"aspect_ratio": "16:9"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Pricing
What it costs on Venice
Pay per clip on Venice — price scales with resolution and duration (5s–15s), from $0.11.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
FAQ
Frequently asked questions
MiniMax H3 Max Turbo is a speed-optimized video generation model that produces 5–15 second clips from a text prompt or a still image, with native synchronized stereo audio, at 480P, 768P, or 1080P across six aspect ratios. It is the distilled, faster tier of the H3 Max family, released in September 2026.
Venice bills per clip, with the price scaling by resolution and duration: $0.11 for a 480P 5-second clip, $0.17 at 768P, and $0.34 at 1080P, rising with clip length up to 15 seconds. There is no subscription requirement — you pay in credits for what you generate.
It is neither free-software nor open-weights: the Turbo checkpoint is proprietary and cannot be self-hosted or fine-tuned. Third-party reporting says the underlying H3 base model is open-weight, but Turbo itself is a closed distilled derivative. On Venice you pay per clip rather than per month.
MiniMax H3 Max Turbo generates at 480P, 768P, and 1080P, in clips from 5 to 15 seconds, across six aspect ratios: 16:9, 21:9, 4:3, 1:1, 3:4, and 9:16. Every clip carries native synchronized audio.
Yes. On Venice the image-to-video variant runs under the model id minimax-h3-max-turbo-image-to-video: you supply a still image and the model animates it into a 5–15 second clip with the same resolution options and native audio as the text-to-video variant. It also accepts first and last frame guidance for controlled transitions.
Both share the same 5–15 second range, 480P–1080P resolutions, and native stereo audio. H3 Max Turbo is roughly twice as fast and cheaper per clip, at the cost of some fine detail in complex shots; H3 Max resolves more micro-detail and suits final renders. A common workflow is iterating on Turbo and re-rendering the winner on H3 Max.
Yes — native stereo audio is predicted together with the picture, not added afterward, so effects, ambience, and dialogue arrive synchronized. Describing the soundscape in your prompt influences what the audio track contains.
Venice runs it under its private, zero-retention tier: prompts and generated clips are not stored, profiled, or used for training, and generation is not tied to a personal history. Note that inference is not end-to-end encrypted or TEE-isolated — the privacy guarantee is zero retention, not encrypted compute.
Clips run from 5 to 15 seconds, selectable in one-second steps, at any of the supported resolutions and aspect ratios. Third-party reviewers suggest 5–10 seconds as the most reliable range for prompt adherence.
Run MiniMax H3 Max Turbo privately
No prompt logging. No data used for training.