Rhythm API · Beta
Rhythm API
Speech recognition from the engine behind Rhythm dictation.
Chinese–English dictation
Chinese, English, or both in one sentence.
BetaStreaming recognition
Partial results while you speak.
Coming in the betaPrivate deployment
Audio never leaves your servers.
On request
Runs on X-ASR zh-en (Apache-2.0) and Whisper large-v3-turbo (MIT)
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["RHYTHM_API_KEY"],
base_url=os.environ["RHYTHM_API_BASE_URL"],
)
with open("audio.wav", "rb") as audio:
transcript = client.audio.transcriptions.create(
model="rhythm-asr-1",
file=audio,
)
print(transcript.text)POST /v1/audio/transcriptions · model rhythm-asr-1Base URL not public yet
How it works
Request access
Tell us what you'll build.
Get a key
Up to 5, once approved.
Call the API
Any OpenAI client.
Pricing
Free during the beta.
After the beta, usage-based pricing per audio hour; we will tell you before any charge.