arbi-tr 0.2.1

Auto-generated Python client for the ARBI-TR transcription API

arbi-tr

Auto-generated Python client for the ARBI-TR transcription API (faster-whisper + pyannote diarization), generated from the service's OpenAPI spec by openapi-python-client.

This package is generated. Do not edit it by hand — run ./scripts/generate-client.sh in the ARBI-TR repo to regenerate it from the current API.

Install

pip install -e client/        # local, from a generated checkout

Usage

Create a client pointed at your ARBI-TR deployment:

from arbi_tr_client import Client

client = Client(base_url="http://localhost:8000")

Async job API (transcribe → poll)

from arbi_tr_client.api.default import (
    transcribe_audio_endpoint_transcribe_post as submit,
    get_task_status_task_status_session_id_get as status,
)
from arbi_tr_client.models import BodyTranscribeAudioEndpointTranscribePost
from arbi_tr_client.types import File

with open("meeting.mp3", "rb") as fh:
    body = BodyTranscribeAudioEndpointTranscribePost(
        file=File(payload=fh, file_name="meeting.mp3", mime_type="audio/mpeg"),
        task_str="transcribe",
    )
    submitted = submit.sync(client=client, body=body)

result = status.sync(client=client, session_id=submitted.session_id)

OpenAI-compatible endpoint

The service also exposes POST /v1/audio/transcriptions, so the standard OpenAI SDK works as a drop-in client. Use transcribe_openai_v1_audio_transcriptions_post from this package for a typed call.