M

Multilingual

MMultilingual
Compare in
on
$0.001per minute
5.4Open ASR
—released
en, es, fr +53Languages
$0.001 / minPer audio minute
batch audio transcription; optional enrichment add-ons billed separatelyPrice covers

Ranking?

Full ranking →
1Azure Speech 07 20261.82Azure Speech 06 20261.93Universal 3.5 Pro2.24Scribe v22.35GPT Transcribe, OpenAI2.3
···
23Phi 4 Multimodal Instruct424Multilingual425Hojo ASR Multi v14

Benchmarks

in Speech & Voice
Open ASR · average WER5.4#12
Open ASR · AMI Cleaned WER6.3#3
Open ASR · Earnings-22 Cleaned AA Chunked WER6#13
Open ASR · GigaSpeech Cleaned WER6.9#2
Open ASR · LibriSpeech Clean WER0.9#1
Open ASR · LibriSpeech Other WER1.8#1
Open ASR · SPGISpeech WER2.1#16
Open ASR · URGENT 2024 WER10#9
Open ASR · Voice Arena Monsoon WER3.3#7
Open ASR · VoxPopuli AA Cleaned WER3.6#27
Open ASR · URGENT 2024 Clean WER4.4#6
Open ASR · French FLEURS WER4.7#21
Open ASR · French MCV WER5.3#5
Open ASR · French MLS WER2.5#3
Open ASR · German FLEURS WER4#24
Open ASR · German MCV WER2.9#8
Open ASR · Spanish FLEURS WER3.8#31
Open ASR · Spanish MCV WER2.8#7
Open ASR · Spanish MLS WER2.4#1
Open ASR · Italian FLEURS WER2.9#24
Open ASR · Italian MCV WER2.5#5
Open ASR · Italian MLS WER4.1#1
Open ASR · Portuguese FLEURS WER3.5#11
Open ASR · Portuguese MLS WER6.3#18
Open ASR · Hindi Monsoon WER28.2#29
Open ASR · Dutch FLEURS WER6.3#19
Open ASR · Dutch MCV WER4.1#12
Open ASR · Dutch MLS WER3.5#1
Open ASR · Armenian FLEURS WER43.8#14
Open ASR · Armenian MCV WER57.4#15

FAQ

How much does Multilingual cost?›

$0.001 per minute.

Is Multilingual free to use?›

There is no free API; it costs $0.001 per minute.

How good is Multilingual, and what is it best at?›

#12 of 61 in Speech to text.

Is Multilingual better than Phi 4 Multimodal Instruct?›

On Open ASR · German FLEURS WER, Multilingual scores 4 vs 4 — about the same. It is 99% cheaper.

Is Multilingual open source?›

No, Multilingual is a closed model available through an API. The closest open-weights option here is Voxtral Small 24B 2507.

What is Multilingual?›

Multilingual is a closed speech & voice model from Modulate.

What are the best alternatives to Multilingual?›

Azure Speech 07 2026 (better), Voxtral Small 24B 2507 (open weights).