Try GPT Image 2.6 Flare by OpenAI from OpenAI →
Models
Agents
Workflows
Studio
PricingBlogDocs
ExploreDiscover models by categoryDiscoverModels and effects built by WiroBrowse All ModelsBrowse the complete catalogSee FavoritesSign in to view saved models
Generative Media AgentCreate and edit media by chattingWorkflow AgentBuild visual workflows with Agent
OverviewThe platform at a glanceLearnSkills, knowledge, guardrailsAnatomyWhat makes agents reasonBuild Your AgentPick connectors and skills, set tier, deploy
Pre-built AgentsBrowse the catalog
Agent Usecases
Ad Campaign ManagerApp Event ManagerApp Review RepliesBarber BookingCustomer Win-BackEcommerce ListingsRestaurant Reviews
Sign InStart Building
Explore models/
Video GenerationActive

MiniMax H3 Highspeed R2V Video

byminimax

Create 4-15s videos from a scene description plus up to 12 reference files. H3 Highspeed R2V keeps characters, style, motion, and audio consistent.

Image to Video
Model ID
h3 highspeed r2v
Provider
minimax
Updated
Nov 5, 2026, 15:01
0
Comments
Average rating : 0 (0 users)
Providerminimax
Modelh3 highspeed r2v
Image to Video
Support Wiro on G2
wiro playground—minimax/h3 highspeed r2v
1 / 9
Maximum 9 image allowed
Delete All
Image 1

Optional. Subject or style references. Up to 9 images. Cite them in the prompt as Image 1, Image 2, and so on. At least one image or video is required when audio is used. Images, videos, and audio together max 12 files.

Required. Describe the video and refer to inputs by order: Image 1, Video 1, Audio 1, and so on.

Required. Length of the video in whole seconds, between 4 and 15.

Required. Default 768P. Choose 480P for a faster, cheaper result.

0 / 3
Maximum 3 video allowed
Delete All

Optional. Motion references. Up to 3 clips, each 2–15 seconds, total at most 15 seconds. Cite them in the prompt as Video 1, Video 2, and so on.

Sample outputs
Updated Nov 5, 2026, 15:01

Overview

MiniMax H3 Highspeed R2V generates short videos from text plus reference assets. It encodes your reference images, video clips, and audio clips, then guides generation to match them. It returns a single MP4 video with synchronized stereo audio, so you can keep the same subject across shots. MiniMax built it as a speed-focused variant of the H3 reference workflow.

What you can build

  • Character-consistent clips from 1–9 reference images
  • Product videos that keep the same item, colors, and materials across motion
  • Style-controlled scenes using multiple reference images as a mood board
  • Motion transfer using a reference clip for pacing and movement
  • Simple video edits by providing a source clip as a reference and describing changes
  • Voice or ambience matching by adding an audio reference alongside visual references

Inputs

  • A scene description written as plain text. Refer to uploaded assets by order, like “Image 1” or “Video 1”.
  • Up to 9 reference images for identity, props, wardrobe, or style. PNG and WEBP are commonly used.
  • Up to 3 reference video clips for motion or editing guidance. MP4 clips are commonly used.
  • Up to 3 reference audio clips for voice or sound cues. MP3 clips are commonly used.
  • If you include audio references, you must also include at least 1 reference image or 1 reference video.
  • A hard cap of 12 total reference files across images, videos, and audio combined.
  • A target clip length in whole seconds, from 4 to 15 seconds.
  • A target output resolution class: 480p or 768p.
  • An aspect ratio selection: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, or 9:16. Adaptive chooses a framing based on your inputs.

Outputs

  • One MP4 video file.
  • Video includes synchronized stereo audio.
  • Audio is 32 kHz stereo.
  • Video runs at 24 fps.
  • Framing follows the requested aspect ratio and the selected resolution class.

Limitations

  • Each generated video is limited to 15 seconds.
  • Reference images are limited to 9 files.
  • Reference videos are limited to 3 clips, and the clips should be short.
  • Reference audio is limited to 3 clips, and audio cannot be the only reference input.
  • Total reference assets are limited to 12 files per run.
  • Output aspect ratios are limited to the supported set listed in the input options.
  • This Highspeed route is intended for 480p or 768p output, not native 2K.
  • Low-quality inputs reduce consistency. Blurry faces, heavy compression, fast cuts, or noisy audio can cause drift.

Safety & compliance

  • Follow MiniMax’s Acceptable Use Policy for the H3 family.
  • Don’t generate unlawful content or content that harms minors.
  • Respect copyright, trademark, and personality rights when using references of people or brands.
  • If you provide this model to end users, add reasonable content safeguards.

Example prompts

Great starting points for h3 highspeed r2v.

Image 1 is the teapot. A hand lifts this exact blue-and-white porcelain teapot and pours steaming golden tea into a small white cup on a sunlit wooden table, steam curling up in the window light, close-up shot. Keep the teapot design identical.Video Generation
Image 1 is the film camera. A young woman with shoulder-length brown hair, a straw hat and a white linen shirt walks through a sunny old-town market with this exact silver and black camera hanging from its brown leather strap, then lifts it to her eye and takes a photo of the colorful fruit stalls. The same woman throughout, the camera following her at chest height. Keep the camera design identical.Video Generation
Image 1 is the man. He walks through a busy night market in the rain wearing the same yellow raincoat, glowing food stalls and lanterns reflected in the wet street, raindrops falling, the camera moving backward in front of him.Video Generation
Image 1 is the toy sailboat. The same wooden toy sailboat, without its display stand, floats on its own hull and glides across a calm park pond on a sunny day, its white sails filling with a light breeze, small ripples spreading behind it, ducks swimming in the distance, low camera at water level. Keep the boat design identical.Video Generation

API quick start

Run h3 highspeed r2v with a single API call.

POST https://api.wiro.ai/v1/Run/minimax/h3-highspeed-r2v
{
  "prompt": "Image 1 is the teapot. A hand lifts this …",
  "inputImage": "https://your-cdn.com/input.png",
  "duration": 5,
  "resolution": "768P"
}
curl
curl -X POST "https://api.wiro.ai/v1/Run/minimax/h3-highspeed-r2v" \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_WIRO_API_KEY" \
  --data-binary @- <<'JSON'
{
  "prompt": "Image 1 is the teapot. A hand lifts this …",
  "inputImage": "https://your-cdn.com/input.png",
  "duration": 5,
  "resolution": "768P"
}
JSON

Other models you might like

Explore similar video generation models.

View all
Image to Video

image-to-video-hailuo-01-subject

minimax

Image to Video

image-to-video-hailuo-01

minimax

Image to Video

image-to-video-hailuo-01-director

minimax

Image to Video

image-to-video-hailuo-02

minimax

  • hailuo-2.3-Fastminimax
  • h3 max r2vminimax
  • h3 lite r2vminimax
  • image-to-video-hailuo-01-liveminimax
  • grok-imagine-r2vxai
  • kling-v2.6-motion-controlklingai
  • Comedy Effectwiro
  • image-to-video-v3.5pixverse

Discover, test, and run AI models, build workflows and agents with one unified API.

All systems operational
WiroModelsAgentsPricingAboutBlogCareersContactPartnerChangelogStatusFAQ
CategoriesText to ImageImage to ImageImage EditingImage to TextText to VideoImage to VideoVideo to VideoText to SpeechSpeech to TextSpeech to SpeechVoice CloneRealtime ConversationMusic GenerationLLMUtilities
ModelsNano Banana 2GPT Image 2.5Seedream V5 ProSeedance 2.5Veo 3.1Kling V3FLUX 3FLUX.2 ProWan 3.0 PrimeGrok Imagine 1.5
PartnersGoogleOpenAIByteDanceBlack Forest LabsKling AIQwenAlibabaxAIMiniMaxElevenLabs
Getting StartedWiro ModelsWiro EffectsIntroductionAuthenticationProjectsCode ExamplesWiro MCP ServerSelf-Hosted MCPn8n IntegrationLLMs.txt
API ReferenceModelsRun a ModelModel ParametersTasksLLM & Chat StreamingWebSocketRealtime VoiceFiles
© 2026 Wiro AI. All rights reserved.
PrivacyTermsData Deletion