Gemini 3.5 Transcribe

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://routerai.ru/api/v1"
)

with open("audio.mp3", "rb") as f:
    transcription = client.audio.transcriptions.create(
        model="google/gemini-3.5-transcribe",
        file=f,
        language="ru"
    )

print(transcription.text)

import requests
import base64
import json

url = "https://routerai.ru/api/v1/audio/transcriptions"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

# Читаем аудиофайл и кодируем в base64
with open("audio.mp3", "rb") as f:
    audio_base64 = base64.b64encode(f.read()).decode("utf-8")

data = {
    "model": "google/gemini-3.5-transcribe",
    "input_audio": {
        "data": audio_base64,
        "format": "mp3"
    },
    "language": "ru"
}

response = requests.post(url, headers=headers, json=data)
result = response.json()
print(json.dumps(result, indent=2, ensure_ascii=False))

# Вариант 1: multipart/form-data (OpenAI-совместимый) — файл как есть
curl -X POST "https://routerai.ru/api/v1/audio/transcriptions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@audio.mp3" \
  -F "model=google/gemini-3.5-transcribe" \
  -F "language=ru"

# Вариант 2: JSON с аудио в base64
curl -X POST "https://routerai.ru/api/v1/audio/transcriptions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-3.5-transcribe",
    "input_audio": {
      "data": "'"$(base64 -w0 audio.mp3)"'",
      "format": "mp3"
    },
    "language": "ru"
  }'