MuST-C

by FBK (Fondazione Bruno Kessler)

TED-based English→X speech translation corpus across 14 target languages.

TL;DR

TED-based English→X speech translation corpus across 14 target languages.

Best for end-to-end speech translation benchmarks (English audio → translated text in 14 languages). Pricing: free.

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

What it is

MuST-C is the largest publicly-available multilingual speech-translation corpus, ~400h per English→target-language pair across 14 targets (de, es, fr, it, nl, pt, ro, ru, ar, cs, fa, tr, vi, zh). License: CC BY-NC-ND 4.0.

Best for: End-to-end speech translation benchmarks (English audio → translated text in 14 languages).
Watch out for: CC BY-NC-ND 4.0 · NON-COMMERCIAL · derived from TED talks; subject to TED licensing · single source language (English). Cite: Di Gangi et al., NAACL 2019; Cattoni et al., 2021.

Install / use

wget https://ict.fbk.eu/must-c/  # request form per release

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeNo
Languages supported14
HIPAA eligibleNo

MuST-C vs Whipscribe

FeatureMuST-CWhipscribe
CategoryOpen sourceTranscription APIs
Pricingfreefree beta
Speaker diarizationNoYes
Word timestampsNoYes
StreamingNoNo
Languages1499
PlatformsWeb, FBKWeb, API, MCP

Alternatives to MuST-C

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.