Speech Revolutions Docs

Production speech-to-text for developers. Transcribe audio with speaker labels, word timestamps, and multiple output formats — via SDK or a single terminal command.

New here? Start with the Quickstart — first transcript in under a minute. Or pick your language:

Python

Sync and async clients, one-line transcribe.

pip install speechrevolutions

JavaScript

TypeScript types, Node and edge runtimes.

npm install speechrevolutions

Context-aware, cancellable on every call.

go get github.com/speechrevolutions/speechrevolutions-go

Async-first client targeting net8.0.

dotnet add package SpeechRevolutions

Choose your path

Use the SDK for apps and pipelines. Use cURL / terminal when you want a single streaming upload without managing upload sessions.

Use case	Endpoint	Best for
SDK / apps	`POST /api/v1/upload`	Large files, progress heartbeats, resumable wait via SSE
Terminal / scripts	`POST /api/v1/transcribe`	One-shot stream upload + percentage progress on the wire

Base URL

All API requests go to https://api.speechrevolutions.com. Authenticate with the X-API-Key header.

30-second taste

from speechrevolutions import SpeechRevolutions

client = SpeechRevolutions()  # SPEECHREVOLUTIONS_API_KEY
result = client.transcribe("meeting.mp3", speaker_labels=True)
print(result.text)

What you get

Zephyr, our speech-to-text engine — ranked #1 on diarization error rate across every subset of our public benchmark suite
Speaker diarization and word-level timestamps
Output formats: txt, json, srt, vtt, docx, pdf
Custom vocabulary / keyterm prompting
Live progress — real-time upload and transcription percentage via SDK console bars or callbacks (something neither AssemblyAI nor Deepgram expose for pre-recorded audio)

Next: Quickstart →