MiniMax H3

MiniMax H3 — premium multimodal tier: first/last frame + image/video/audio references, native audio, 2K (default) or 768P output, 4-15s per-second pricing.

MiniMax H3 examples

Credits

Use MiniMax H3 in Nodaro

Add one of these nodes to your canvas, open its settings, and choose MiniMax H3 as the model:

• Generate Video
• Generate Video Pro
• Lip Sync

Run MiniMax H3 from code

Each call starts a job and returns its jobId. Poll GET /v1/jobs/:id/status, or let the SDK wait for the result. A run costs the credits on this page. API reference

With Nodaro connected to Claude, ChatGPT or Cursor, ask in plain words, or have the assistant call the tool: “Use MiniMax H3 to make: A paper boat drifts down a rain-soaked street at dusk, slow dolly in”

curl -s -X POST https://app.nodaro.ai/v1/text-to-video \  -H "Authorization: Bearer $NODARO_API_KEY" \  -H "Content-Type: application/json" \  -d '{        "prompt": "A paper boat drifts down a rain-soaked street at dusk, slow dolly in",        "provider": "minimax-h3",        "duration": 4      }'

import { createClient, StaticTokenAuth } from "@nodaro/sdk"const client = createClient({  baseUrl: "https://app.nodaro.ai",  auth: new StaticTokenAuth(process.env.NODARO_API_KEY!),})const output = await client.nodes.runAndWait("text-to-video", {  prompt: "A paper boat drifts down a rain-soaked street at dusk, slow dolly in",  provider: "minimax-h3",  duration: 4,})

{  "tool": "generate_video",  "arguments": {    "prompt": "A paper boat drifts down a rain-soaked street at dusk, slow dolly in",    "model": "minimax-h3",    "duration": 4  }}

Prompt tips

• Natural-language prompts up to 7000 chars. Front-load what matters: subject → action → scene/environment → lighting → camera move → style. One camera movement per shot.
• Output 2K (default) or 768P (cheaper tier). Aspect: pure text-to-video needs a concrete ratio (21:9/16:9/4:3/1:1/3:4/9:16); reference runs default to adaptive (match the input).
• References go by ordinal in attachment order (@Image 1, Video 1) — earlier = higher priority. Caps: 9 images, 3 videos (2-15s each, ≤15s total), 3 audio clips (≤15s total).
• Reference audio drives speech/lip-sync but never rides alone — pair it with an image or video reference. Audio input is free; the first 5 input images are free, extras bill per image.
• No negative-prompt parameter — put constraints in the prompt text: 'keep it subtitle-free, do not generate a watermark, do not generate a logo'.

Similar models