Typecast api with quick cloning (python)

typecast-quick-cloning.py

import os
import time
import requests


API_TOKEN = {{your API token}}
REFERENCE_AUDIO_FILENAME = {{your audio filename}}


HEADERS = {'Authorization': f'Bearer {API_TOKEN}'}

_, reference_audio_format = os.path.splitext(REFERENCE_AUDIO_FILENAME)
reference_audio_format = reference_audio_format[1:]

# 1. request upload url
r = requests.post(
    "https://typecast.ai/api/speak/resource",
    headers=HEADERS,
    json={"audio_format": reference_audio_format},
)

result = r.json()["result"]
speak_resource_id = result["speak_resource_id"]
presigned_url = result["presigned_url"]

# 2. upload audio file
with open(REFERENCE_AUDIO_FILENAME, "rb") as f:
    requests.put(presigned_url, data=f.read())

# 3. request speech
r = requests.post(
    "https://typecast.ai/api/speak",
    headers=HEADERS,
    json={
        "tts_mode": "audio_file",
        "speak_resource_id": speak_resource_id,
        "text": "Hello world!",
        "lang": "en-us",
        "xapi_hd": True,
    },
)
speak_url = r.json()["result"]["speak_v2_url"]

# 4. polling the speech result
for _ in range(120):
    r = requests.get(speak_url, headers=HEADERS)
    ret = r.json()['result']
    # audio is ready
    if ret['status'] == 'done':
        # download audio file
        r = requests.get(ret['audio_download_url'])
        with open('out.wav', 'wb') as f:
            f.write(r.content)
        print(f"finish")
        break
    else:
        print(f"status: {ret['status']}, waiting 1 second")
        time.sleep(1)