Google Speech Research

by Google Research

Google Research Speech — USM + Chirp + AudioPaLM + FLEURS origins.

TL;DR

Google Research Speech — USM + Chirp + AudioPaLM + FLEURS origins.

Best for frontier multilingual ASR (USM/Chirp 300 languages); speech-LLM hybrids (AudioPaLM). Pricing: free.

Category
Open source
License
—
Stars
—
Last push
—
Pricing
free
Platforms
Web

What it is

Google Speech Research built USM (1000 languages target), Chirp (Vertex AI), AudioPaLM, and shipped FLEURS / PRESTO / SpeechStew benchmarks. Frontier multilingual + speech-LM research.

Best for: Frontier multilingual ASR (USM/Chirp 300 languages); speech-LLM hybrids (AudioPaLM).
Watch out for: Models are proprietary on Vertex AI (Chirp); datasets (FLEURS, PRESTO) open CC BY 4.0.

Install / use

https://research.google/teams/speech/

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeNo
Languages supported300
HIPAA eligibleNo

Google Speech Research vs Whipscribe

FeatureGoogle Speech ResearchWhipscribe
CategoryOpen sourceTranscription APIs
Pricingfreefree beta
Speaker diarizationNoYes
Word timestampsNoYes
StreamingNoNo
Languages30099
PlatformsWebWeb, API, MCP

Alternatives to Google Speech Research

OpenAI Whisper
OpenAI

The reference open-source multilingual ASR model from OpenAI.

OSS · MIT ★ 98.1k
whisper.cpp
Georgi Gerganov

C/C++ port of Whisper — runs on anything, from a Raspberry Pi to Apple Silicon.

OSS · MIT ★ 48.8k
faster-whisper
SYSTRAN

4× faster than reference Whisper using CTranslate2 — production sweet spot.

OSS · MIT ★ 22.3k

Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.