> ## Documentation Index
> Fetch the complete documentation index at: https://docs.perchanceai.art/agent-docs/llms.txt
> Use this file to discover all available pages before exploring further.

# MiniMax H3 API Documentation

> Prompt or up to 4 images + 3 audio clips · Up to 10s · 1080p

POST https://api.perchanceai.art/api/v1/developer/generations

## Model

- `model_key`: `minimax-h3`
- `modality`: `video`
- `operation`: `video_generation`
- price: 20 credits per second

## Authorizations

| Name | Type | In | Required | Description |
| --- | --- | --- | --- | --- |
| `Authorization` | string | header | required | Formal account API Key or OAuth token, for example `Bearer pca_live_...` |
| `Idempotency-Key` | string | header | required | Unique key for the generation request |
| `X-App-Id` | string | header | optional | Fixed to `perchanceai` for this deployment |

## Body

`application/json`

| Field | Type | Required |
| --- | --- | --- |
| `negative_prompt` | string | optional |
| `prompt` | string | optional |
| `reference_upload_ids` | array | optional |

## Parameters

| Name | Accepted values |
| --- | --- |
| `supports_reference` | true |
| `video_modes` | t2v, r2v, i2v |
| `resolutions` | 480p, 768p, 1080p |
| `max_prompt_chars` | 8000 |
| `aspect_ratios` | 16:9, 9:16, 1:1 |
| `video_input_limits` | [object Object] |
| `supports_first_frame` | true |
| `supports_multi_reference` | true |
| `supports_last_frame` | true |
| `sizes` | 1024x1792, 1792x1024, 1024x1024, 720x1280, 1280x720 |
| `duration_seconds` | 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
| `max_reference_uploads` | 8 |
| `max_n` | 1 |
| `resolution_names` | 480p, 768p, 1080p |
| `presets` | fun, normal, spicy, custom |
| `reference_image_count` | 4 |
| `max_total_reference_bytes` | 41943040 |

## Example request

```bash
curl --request POST \
  --url https://api.perchanceai.art/api/v1/developer/generations \
  --header 'Authorization: Bearer <pca_live_key>' \
  --header 'X-App-Id: perchanceai' \
  --header 'Content-Type: application/json' \
  --header 'Idempotency-Key: <uuid>' \
  --data '{"model_key":"minimax-h3","operation":"video_generation","prompt":"your prompt"}'
```

## Responses

- `202` — accepted; durable generation identifier returned, not task completion.
- `401` `DEVELOPER_TOKEN_REQUIRED` — the bearer token is missing.
- `401` `DEVELOPER_TOKEN_INVALID` — the bearer token is invalid, including historical anonymous credentials.
- `422` `FIELD_INVALID` / `FIELD_REQUIRED` — the payload does not match the model schema.
- `429` `RATE_LIMITED` — too many requests.