MiniMax · Hailuo
Shanghai lab founded by ex-SenseTime researchers, and one of the few to open-weight its flagship LLMs. Hailuo delivers cinematic video; the M-series ships streaming coding and agent models.
Models
Text-to-image generation and multi-reference editing at up to 20 megapixels
Faster H3 Image generation and multi-reference editing at up to 20 megapixels
Fast H3 video generation for everyday content, with first-to-last frame control and multimodal references
Lightweight H3 video generation for previews, drafts and batch variations, with frame and reference control
MiniMax M3.1 Flash
compatible APIEfficient open-weight multimodal model for coding, agents, and 1M-token context
Fast multimodal video generation with native audio and image, video, and audio reference control
Distilled H3 Max video generation for faster iteration with native audio and first-to-last frame control
High-throughput video generation with stronger prompt adherence, native synced audio, and first-to-last frame control
Multimodal video generation with native synced audio, multi-reference consistency, and continuation workflows
Frontier multimodal agentic model with 1M context for coding, long-horizon tasks, and visual understanding
MiniMax M2.7
deprecatedLong-context agentic coding and office productivity model for fast, reliable tool use
MiniMax M2.7 Highspeed
deprecatedFaster throughput for agentic coding and tool-driven automation
Model guides
Prompting
How to write text-to-image prompts for MiniMax H3 Image: sizing by preset, exact dimensions or no size, and planning the task and layout before the details.
Read the guide →MiniMax H3 ImagePrompt optimizer
How to use the prompt optimizer on MiniMax H3 Image: what the rewrite adds to a short prompt, and when to turn it off and send the prompt as written.
Read the guide →MiniMax H3 ImageText and layouts
How to get exact copy and structured layouts from MiniMax H3 Image: quoted text, charts and figures, dense pages, app screens, and grids and storyboards.
Read the guide →MiniMax H3 ImageEditing images
How to edit one image with MiniMax H3 Image: naming the target in words, moving, copying, replacing and deleting objects, rewriting text, and editing over several rounds.
Read the guide →MiniMax H3 ImageMulti-reference composition
How to compose one image from up to 16 references with MiniMax H3 Image: giving each image a role, borrowing color, following a layout sketch and furnishing a room.
Read the guide →MiniMax H3 MaxGenerating video
How to generate video with MiniMax H3 Max: the fixed size pairs at 480p and 768p, animating a first and last frame, native synced audio, and rendering faster than playback.
Read the guide →MiniMax H3Generating video
How to generate video from text with MiniMax H3: the six 2K aspect ratios, 5 to 15 second durations, native synced audio, and prompting for cinematic shots.
Read the guide →MiniMax H3Motion, camera, and performance
How to transfer motion, a camera move, or an acting performance from a reference video onto a new subject with MiniMax H3 Omni Reference.
Read the guide →MiniMax H3Sound and voice
How to direct MiniMax H3's native audio from the prompt: ambience, synced sound effects, music, and spoken dialogue with lip-sync.
Read the guide →MiniMax H3First and last frame
How to animate a still image with MiniMax H3 and bridge a first and last frame into one continuous shot, with the output following the image's own aspect ratio.
Read the guide →MiniMax H3Reference-driven consistency
How to lock a character, product, or style across a new MiniMax H3 shot with Omni Reference images, and address each reference by index in the prompt.
Read the guide →MiniMax H3Editing video
How to edit a finished clip with a prompt in MiniMax H3: replace a subject, relight a scene, add or remove elements, and combine edits, keeping the rest untouched.
Read the guide →Related collections
Browse all collections→FLUX
Black Forest Labs flagship family. Strong prompt fidelity and consistency, with lighter Klein variants.
Open →Wan
Alibaba's multimodal family. Strong video and image generation, with open-weights variants for self-hosting.
Open →Kling
High-end video generation with strong motion coherence. Production-grade 4K variants, image-to-video pipelines, and avatar generation.
Open →Veo
Google's frontier text-to-video family. Cinematic motion, native audio synthesis, and a premium production tier built for ads and short-form.
Open →