---
title: "Generate a video (minimax-h3-spicy)"
method: POST
path: "/v1/minimax-h3-spicy"
tags: ["Tasks"]
---

# Generate a video (minimax-h3-spicy)

`POST /v1/minimax-h3-spicy`

## Request

- Base URL: `https://api.hotapi.ai`
- URL: `https://api.hotapi.ai/v1/minimax-h3-spicy`
- Auth: HTTP bearer

Uncensored video generation from text, a first-frame image, first+last frames, or omni-reference (up to 4 reference images and/or 1 reference video) — 480p/768p/1080p/2k/4k, 5–15s, selectable aspect ratio, generated audio included. Async: returns a task (202); poll `GET /v1/tasks/{id}`.

## Query parameters

- `webhook` string, uri

## Headers

- `idempotency-key` string

## Request body

- object: Uncensored text / first-frame / first+last-frame / omni-reference video input (MiniMax H3). Output includes a generated audio track.
  - `prompt` string, required: Text prompt (max 2000 characters).
  - `image_url` string: Optional first-frame image (HTTPS URL or base64). When present the task runs image-to-video; omitted = text-to-video. Cannot be combined with reference fields.
  - `last_frame_url` string: Optional last-frame image (HTTPS URL or base64) for first-to-last-frame interpolation. Requires image_url; cannot be combined with reference fields.
  - `reference_image_urls` string[]: 1–4 reference images (HTTPS URLs or base64) for omni-reference generation — the output carries their subject/style. Cannot be combined with image_url/last_frame_url. Billed output-only, same as text-only.
  - `reference_video_urls` string[]: At most 1 HTTPS reference video URL (0.5–600s, ≤100MB) for omni-reference generation. Cannot be combined with image_url/last_frame_url. Billed output-only — the reference video's own duration is NOT charged.
  - `duration_seconds` integer, required: 5–15 seconds.
  - `resolution` union: Output resolution; omitted = cheapest tier (480p).
    - '480p'
    - '768p'
    - '1080p'
    - '2k'
    - '4k'
  - `ratio` union: Output aspect ratio.
    - '16:9'
    - '9:16'
    - '1:1'
  - `seed` integer, nullable: Random seed (0–2147483647). null or omitted = random.

## Response `200`

An async task. Creation returns 202 with status=queued; poll GET /v1/tasks/{id} (or await the webhook) until a terminal status.

- Task: An async task. Creation returns 202 with status=queued; poll GET /v1/tasks/{id} (or await the webhook) until a terminal status.
  - `id` string, required
  - `object` 'task', required
  - `status` union, required: Lifecycle status. Terminal states: succeeded | failed | cancelled | expired.
    - 'pending'
    - 'queued'
    - 'processing'
    - 'succeeded'
    - 'failed'
    - 'cancelled'
    - 'expired'
  - `model` string, required
  - `task_kind` string, required
  - `input` object, required: The normalized request input the task was created with.
  - `output` object: Result payload, present once status=succeeded. Contains hosted asset URLs (re-hosted on our CDN).
  - `error` object
    - `type` string, required
    - `code` string, required
    - `message` string, required
  - `estimated_credits_cost` integer, required: Credits charged up-front when the task was created (1000 credits = $1).
  - `reference_video_seconds` number: Input-video seconds measured and frozen at submission; reused for draft finalization billing. Zero when no reference video was supplied.
  - `actual_credits_cost` integer: Final credits charged after settlement (equals estimate today).
  - `webhook_url` string, uri
  - `api_key_id` string: The organization's API key that created the task; absent for dashboard/session-created tasks.
  - `created_at` integer, required: Unix seconds.
  - `queued_at` integer: Unix seconds.
  - `started_at` integer: Unix seconds.
  - `completed_at` integer: Unix seconds.
  - `expires_at` integer, required: Unix seconds; result + asset retention deadline.

## Other responses

- `202`: An async task. Creation returns 202 with status=queued; poll GET /v1/tasks/{id} (or await the webhook) until a terminal status.
- `400`: Standard error envelope returned by every error response (OpenAI-style).
- `401`: Standard error envelope returned by every error response (OpenAI-style).
- `402`: Standard error envelope returned by every error response (OpenAI-style).
- `404`: Standard error envelope returned by every error response (OpenAI-style).
- `429`: Standard error envelope returned by every error response (OpenAI-style).

---

[API](https://skmtc.dev/hotapi/apis/hotapi-gateway.md) · [All operations](https://skmtc.dev/hotapi/apis/hotapi-gateway/llms.txt) · [OpenAPI document](https://skmtc.dev/hotapi/apis/hotapi-gateway/revisions/9b6578b0c389?raw)