Malay speech-to-text API: 8.6% word error, measured

Measured accuracy

Model	Word error rate	Character error rate	Clips with no errors	Empty transcripts	Speed
Whisper Large v3 `openai/whisper-large-v3`	8.6%	3.2%	10 of 30	0 of 30	0.067× real time

Good. Usable with light review. Expect to correct a word here and there, mostly names and numbers. The clips are clean, single-speaker read speech, so treat this as a best case for your own recordings. How we measured, and all 83 languages.

What a transcript looks like

The clip with the median error rate from the test, reference text against what the API returned.

Spoken	Gangguan tidur ialah proses bangun secara sengaja selama tempoh tidur normal anda dan tertidur sekejap (10-60 minit).
Whisper Large v3	Gangguan tidur ialah proses bangun secara sengaja selama tempoh tidur normal anda dan tertidur sekejap 10 ke 60 minit.

Language detection

Send language=ms. Of 10 clips sent without it, only 7 were identified as Malay; the others were labelled Indonesian (3). When detection is wrong, Whisper writes the transcript in the language it guessed.

Quickstart

The request is the OpenAI transcription request with a different base URL. language is optional in the API but, as shown above, not optional in practice for Malay.

cURL

curl https://api.inferenceapis.com/v1/audio/transcriptions \
  -H "Authorization: Bearer $INFERENCE_API_KEY" \
  -F "model=openai/whisper-large-v3" \
  -F "language=ms" \
  -F "file=@recording.mp3" \
  -F "response_format=verbose_json"

Python · openai SDK

import os
from openai import OpenAI

client = OpenAI(base_url="https://api.inferenceapis.com/v1", api_key=os.environ["INFERENCE_API_KEY"])
with open("recording.mp3", "rb") as f:
    r = client.audio.transcriptions.create(model="openai/whisper-large-v3", file=f, language="ms", response_format="verbose_json")
print(r.text)

Subtitles and timestamps

response_format=srt or vtt returns a subtitle file with cues of up to two lines, broken at sentence ends and pauses. verbose_json with timestamp_granularities[]=word returns a start and end time for every word if you want to build your own.

Cost

$0.002 per minute of audio on either model: $0.12 for an hour, $12 for a hundred hours. Duration is measured from the file you upload. There is no daily cap on minutes and no subscription; enough to run your own Malay test.

Notes for Malay

Malay and Indonesian (id) are easily confused by auto-detection, as the result above shows.

FAQ

How accurate is Whisper for Malay?

On 30 clean read-speech clips it made 8.6% word errors, about one word in 12. Usable with light review. Expect to correct a word here and there, mostly names and numbers. Real recordings with noise, accents or several speakers will score worse.

Does language auto-detection work for Malay?

Not reliably. Of 10 clips sent without a language, 7 were identified as Malay; the rest were labelled Indonesian (3). A wrong label means the transcript is written in the wrong language, so always send language=ms.

Can it translate Malay speech to English?

The API returns Malay text. For English, transcribe first and translate the text with a chat model on the same key.

Related languages: Indonesian (5.5%) · Javanese (71%)