Tacotron 2

by Google / NVIDIA reference

Google's seminal end-to-end TTS architecture — the neural-TTS starting point.

TL;DR

Google's seminal end-to-end TTS architecture — the neural-TTS starting point.

Best for education and academic baselines; legacy systems still using Tacotron-style pipelines. Pricing: free (BSD-3-Clause).

Category
Open source
License
—
Stars
—
Last push
—
Pricing
free (BSD-3-Clause)
Platforms
Linux

What it is

Tacotron 2 is the 2017 Google paper that defined modern end-to-end neural TTS. NVIDIA's BSD-licensed PyTorch implementation is the most-used public reference. Consent posture: synthetic-only architecture.

Best for: Education and academic baselines; legacy systems still using Tacotron-style pipelines.
Watch out for: Slower and lower-quality than VITS / XTTS; effectively English-only in the public checkpoint.

Install / use

git clone https://github.com/NVIDIA/tacotron2

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeNo
Languages supported1
HIPAA eligibleNo

Tacotron 2 vs Whipscribe

FeatureTacotron 2Whipscribe
CategoryOpen sourceTranscription APIs
Pricingfree (BSD-3-Clause)free beta
Speaker diarization—Yes
Word timestamps—Yes
Streaming—No
Languages199
PlatformsLinuxWeb, API, MCP

Alternatives to Tacotron 2

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.