# Fantase API

> AI image and video generation API. REST endpoints with API-KEY header auth. Supports image-to-image, image-to-video, and image-to-audio+video generation.

Base URL: https://api.fantaseai.com

## Authentication

All requests require an `API-KEY` header. Get your key at https://fantaseai.com/dashboard.

## API Reference

- [OpenAPI Spec (JSON)](https://api.fantaseai.com/api/openapi.json): Full machine-readable API schema. Requires API-KEY header.
- [Interactive Docs (Swagger UI)](https://api.fantaseai.com/api/docs): Try endpoints in the browser.

## Endpoints

### POST /api/create_image

Transform an image using AI. Returns a Task object.

Request body:
- `input_image_base64` (string, required): Base64-encoded input image.
- `preset_name` (string, required): Transformation preset. Examples: `custom`, `tattoo`, `restyle`. See full list in OpenAPI spec.
- `prompt` (string, optional): Custom prompt text. Used with `custom` preset or to override preset prompt.
- `input_image2_base64` (string, optional): Second image for remix presets (e.g. `remix_dress`, `remix_yoga`).
- `expand_prompt` (boolean, optional): Enable LLM-based prompt upsampling.

### POST /api/create_video

Generate video from a single image. Returns a Task object.

Request body:
- `input_image_base64` (string, required): Base64-encoded input image.
- `preset_name` (string, required): Motion preset. Examples: `custom`. See full list in OpenAPI spec.
- `prompt` (string, optional): Custom prompt text.
- `quality` (string, optional): `low`, `mid`, or `high`. Default: `low`.

### POST /api/create_audio_video

Generate video with synchronized audio. Returns a Task object.

Request body:
- `input_image_base64` (string, required): Base64-encoded input image.
- `preset_name` (string, required): Preset. See valid values in OpenAPI spec.
- `prompt` (string, optional): Custom prompt text.
- `quality` (string, optional): `low` or `mid`. Default: `low`.
- `video_length` (string, optional): `5s`, `10s`, `15s`, or `20s`. Default: `5s`.

### GET /api/task/{task_id}

Poll task status and retrieve output URL.

Response fields:
- `id` (string): Task ID.
- `status` (string): `NEW`, `PROCESSING`, `COMPLETED`, `FAILED`, or `DOWNLOADED`.
- `output_url` (string): Direct download URL when status is `COMPLETED` or `DOWNLOADED`.
- `task_message` (string): Error message if status is `FAILED`.
- `cost_amount` (string): USD cost charged for this task.

### GET /api/balance

Check remaining USD balance.

Response fields:
- `symbol` (string): Always `USD`.
- `available_amount` (string): Available balance.
- `frozen_amount` (string): Amount frozen for in-progress tasks.

## Quickstart

```python
import requests, base64, time

API_KEY = "your_api_key"
HEADERS = {"API-KEY": API_KEY}

# 1. Encode image
with open("input.jpg", "rb") as f:
    img_b64 = base64.b64encode(f.read()).decode()

# 2. Create task
resp = requests.post(
    "https://api.fantaseai.com/api/create_video",
    headers=HEADERS,
    json={"input_image_base64": img_b64, "preset_name": "custom", "quality": "high"}
)
task_id = resp.json()["id"]

# 3. Poll until complete
while True:
    result = requests.get(f"https://api.fantaseai.com/api/task/{task_id}", headers=HEADERS).json()
    if result["status"] in ("COMPLETED", "DOWNLOADED"):
        print(result["output_url"])
        break
    elif result["status"] == "FAILED":
        print("Failed:", result["task_message"])
        break
    time.sleep(3)
```

## Resources

- [API Console / Dashboard](https://fantaseai.com/dashboard): Manage API key, balance, and deposits.
- [OpenAPI Spec](https://api.fantaseai.com/api/openapi.json): Machine-readable schema (requires API-KEY header).

## Optional

- [Support (Telegram)](https://telegram.me/Michael_Voss): Volume discounts, custom pricing, technical help.