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

# Python

> Call the Context Aware ASR transcription API from Python.

Install `requests` if needed:

```bash theme={null}
pip install requests
```

## Basic request

```python theme={null}
import json
import requests

messages = [
    {
        "role": "system",
        "content": "You are helping inside a Python and TypeScript voice agent repo.",
    },
    {
        "role": "user",
        "content": "The transcript may include LangGraph, QdrantVectorCache, RAGPipelineConfig, and api.voiceos.dev.",
    },
]

with open("sample.mp3", "rb") as audio:
    response = requests.post(
        "https://beta.api.voiceos.com/v1/audio/transcriptions",
        files={"file": ("sample.mp3", audio, "audio/mpeg")},
        data={
            "messages": json.dumps(messages),
            "languages": json.dumps(["en", "ja"]),
            "dictionary": json.dumps(["VoiceOS", "ASR"]),
            "response_format": "json",
        },
        timeout=60,
    )

response.raise_for_status()
result = response.json()

print("Transcript:", result["text"])
```

## Error handling

```python theme={null}
try:
    response.raise_for_status()
except requests.HTTPError:
    print(response.status_code)
    print(response.json())
    raise
```