# MiniMax H3 Max API on MLMux

> MiniMax H3 Max, the highest-quality H3 model, at 480p or 768p.

- Model page: https://www.mlmux.com/minimax/h3-max
- API model ID: `minimax/h3-max`
- Provider: MiniMax
- Generation modes: Text to video, Image to video, First & last frame, Multimodal reference
- Endpoint: `POST https://api.mlmux.com/v1/videos/generations`
- Reference images: up to 2
- Resolutions: 480p, 768p
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
- Duration: 5 to 15 seconds

## Pricing

Pay as you go in credits; 1 credit = 1 US dollar, with no subscription. `official-cheap` is the default route.

- 480p: $0.05 on `official`, $0.045 on `official-cheap` per second of output video; with a reference video, $0.055 on `official`, $0.05 on `official-cheap`
- 768p: $0.08 on `official`, $0.072 on `official-cheap` per second of output video; with a reference video, $0.143 on `official`, $0.129 on `official-cheap`
- Reference videos are billed by their length too: the charge covers the output seconds plus the reference video's seconds, at the same rate.

## Request

Send your API key as `Authorization: Bearer $MLMUX_API_KEY`. Create keys at https://www.mlmux.com/console/api-keys.

```bash
curl https://api.mlmux.com/v1/videos/generations \
  -H "Authorization: Bearer $MLMUX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "minimax/h3-max",
  "route_mode": "official-cheap",
  "prompt": "A red fox walks through fresh snow at sunrise, slow tracking shot, cinematic light",
  "mode": "text_to_video",
  "ratio": "16:9",
  "resolution": "768p",
  "duration": 5
}'

# The response is a task. Poll it until "status" is "succeeded", then read "video".
curl https://api.mlmux.com/v1/videos/generations/TASK_ID \
  -H "Authorization: Bearer $MLMUX_API_KEY"
```

```python
import os
import time
import requests

headers = {"Authorization": f"Bearer {os.environ['MLMUX_API_KEY']}"}
response = requests.post(
    "https://api.mlmux.com/v1/videos/generations",
    headers=headers,
    json={
      "model": "minimax/h3-max",
      "route_mode": "official-cheap",
      "prompt": "A red fox walks through fresh snow at sunrise, slow tracking shot, cinematic light",
      "mode": "text_to_video",
      "ratio": "16:9",
      "resolution": "768p",
      "duration": 5
    },
    timeout=300,
)
response.raise_for_status()
task = response.json()

while task["status"] not in ("succeeded", "failed", "expired", "canceled"):
    time.sleep(5)
    task = requests.get("https://api.mlmux.com" + task["poll_url"], headers=headers, timeout=30).json()

print(task["status"], task.get("video"))
```

The response is a task. Poll its `poll_url` every few seconds until `status` is `succeeded`, then read the video URL, or receive a webhook when it finishes.

## Frequently asked questions

### What is MiniMax H3 Max?

MiniMax H3 Max is a video generation model from MiniMax. On MLMux you can run it in the browser or call it through one API. MiniMax H3 Max, the highest-quality H3 model, at 480p or 768p.

### How do I use the MiniMax H3 Max API?

Create an API key in the MLMux Console, then send a POST request to https://api.mlmux.com/v1/videos/generations with the key in the Authorization header and model set to minimax/h3-max. The request returns a task: poll it, or receive a webhook when the video is ready.

### What can MiniMax H3 Max do?

Generation modes: Text to video, Image to video, First & last frame, and Multimodal reference. Clips run 5 to 15 seconds at 480p or 768p, in aspect ratios such as 16:9, 9:16, 1:1, 4:3, 3:4, and 21:9, with up to 2 reference images.

### How much does MiniMax H3 Max cost?

MLMux is pay as you go, with no subscription or monthly fee, and 1 credit equals 1 US dollar. One second of 768p MiniMax H3 Max video costs $0.072 on the default official-cheap route and $0.08 on the official route. Rates for every resolution are listed in the pricing section of this page.

### Can I try MiniMax H3 Max without writing code?

Yes. Sign in and use the playground at the top of this page to run MiniMax H3 Max with the same settings as the API. The playground's code views show the matching request for your settings, and the quick start below has copy-paste examples in cURL, Python, and JavaScript.

## More

- Video API guide: https://www.mlmux.com/docs/videos
- Pricing for every model: https://www.mlmux.com/pricing
- All models: https://www.mlmux.com/llms.txt