Choose a category
3 models available
8 models available
3 models available
5 models available
4 models available
8 models available
Create 4–15 second scenes with native audio at 768P or 2K.
Describe the subject, action, camera movement and sound. Include dialogue in quotation marks.
Choose 4–15 seconds. Longer clips use more credits.
Start at 768P; choose 2K for a higher-resolution output.
Turn a written idea into a short video with MiniMax H3. Describe the scene, action, camera and sound, then choose 4–15 seconds at 768P or 2K. This text-to-video workflow starts without uploaded media and includes native generated audio.
Start with one scene, one camera move and a clear sound direction.
Examples for 5 seconds from text, including generated audio. The current price depends on duration and resolution; these image inputs have no separate surcharge.
| Resolution | Duration | Flow credits |
|---|---|---|
| 768P | 5s | 9 |
| 2K | 5s | 15 |
Credits are held on submission and released for confirmed failed generations. Compare credit plans and other models.
Choose the controls that fit your source material.
A prompt can describe a subject, setting, movement, dialogue and ambient sound together. For a cafe scene, specify the camera moving toward a cup, rain outside and quiet room tone. Audio is generated with the video; this page has no separate audio toggle.
Use 16:9 for a landscape player, 9:16 for a vertical clip or 1:1 for a square composition. H3 also offers 21:9, 4:3 and 3:4. Start with the intended framing so important subjects do not rely on a later crop.
Both resolution settings accept integer durations from 4 to 15 seconds. The default is a 5-second 768P clip. Increasing duration or resolution increases the credit cost, which you can review before submitting.
Stay here when you want to create from words alone. Image to Video uses an opening or closing frame. Reference to Video uses uploaded images as visual guidance for a newly composed scene. Each workflow has its own input controls.
Change one part of the input at a time so you can judge what helped.
Prepare the input, describe the change, then review the credit cost.
Name the subject and surroundings, then describe what changes. For example: “A small glass greenhouse in a rain-soaked garden. The camera slowly pushes in as warm lights turn on inside.”
Give the scene a pace and a sound direction: “Gentle rain on glass, soft wind in the leaves, no dialogue.” If speech matters, put the exact line in quotation marks and identify the speaker.
Choose duration, resolution and aspect ratio. Check the displayed credits, sign in and submit. Your task appears in history; refine one part of the prompt between attempts so you can judge what changed.
Change one part of the input at a time so you can judge what helped.
An original H3 text-to-video result generated on Flow Veo 3. Play the video to hear the generated audio. No images were supplied.
Prompt: An elegant glass greenhouse at sunrise, a single orange butterfly rests on a green leaf in the foreground. The camera slowly pushes closer as the butterfly opens its wings and lifts into warm golden light. Dew sparkles, realistic natural textures, cinematic macro photography. Gentle birdsong and soft leaf rustling, no music, no speech, no text or logos.
Answers about this workflow, inputs and output settings.
MiniMax H3 is a video generation model with native audio. This page runs its text-to-video workflow on Flow Veo 3, an independent service that is not affiliated with MiniMax. Some services also call the model Hailuo 03 or Hailuo 3.0.
Yes. This workflow takes a text prompt and output settings. For a specific product, person or composition that you want to supply visually, switch to the image or reference-image workflow.
Describe the intended sound in your prompt, including dialogue in quotation marks where needed. H3 generates audio alongside the video. Exact pronunciation, timing and musical results can vary, so review the output.
The Hailuo 2.3 pages on this site animate an input image without generated audio. H3 adds a dedicated text workflow, native audio, first/last-frame controls and a reference-image workflow, with different output settings and prices.
Opening the tool is free, but generation spends Flow Veo 3 credits. The displayed cost depends on duration and resolution. Check your balance and the pricing page before submitting; do not assume a signup gift covers a run.
Credits are held when a task is submitted. A confirmed failed generation releases that hold. A temporary loss of status updates does not by itself mean the task failed; check history again for its final state.