Amphion

by Open Multimedia AI Lab

Open-source toolkit for audio, music, and speech generation.

TL;DR

Open-source toolkit for audio, music, and speech generation.

Best for research teams running multi-task audio-generation experiments — TTS, singing, vocoder, source separation. Pricing: free (MIT).

Category
Open source
License
—
Stars
—
Last push
—
Pricing
free (MIT)
Platforms
Linux

What it is

Amphion is a research-grade toolkit covering TTS, singing-voice synthesis, voice conversion, and audio generation. MIT-licensed. Used as a benchmark harness for new architectures. Consent posture: synthetic-only in default recipes.

Best for: Research teams running multi-task audio-generation experiments — TTS, singing, vocoder, source separation.
Watch out for: Heavy framework; primarily a research harness rather than a production library.

Install / use

git clone https://github.com/open-mmlab/Amphion

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeNo
Languages supported10
HIPAA eligibleNo

Amphion vs Whipscribe

FeatureAmphionWhipscribe
CategoryOpen sourceTranscription APIs
Pricingfree (MIT)free beta
Speaker diarization—Yes
Word timestamps—Yes
Streaming—No
Languages1099
PlatformsLinuxWeb, API, MCP

Alternatives to Amphion

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.