Afrikaans speech-to-text
FairWe sent 30 Afrikaans test clips through this API. Whisper Large v3 made 30% word errors, about one word in 3. $0.002 per audio minute, word timestamps, SRT and VTT subtitles, files up to 100 MB.
Measured accuracy
| Model | Word error rate | Character error rate | Clips with no errors | Empty transcripts | Speed |
|---|---|---|---|---|---|
Whisper Large v3openai/whisper-large-v3 | 30% | 9.9% | 0 of 30 | 0 of 30 | 0.061× real time |
Fair. Gets the content across, but needs a human pass before it is published or quoted. 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 | n Paar weke gelede, nadat die informasie deur die joernalis Makis Triantafylopoulos in sy gewilde Televisieprogram "Zoungla" op Alfa TV gepubliseer is, is 'n lid van die parlement, en prokureur, Petros Mantouvalos, uit sy posisie onthef, aangesien lede van sy kabinet betrokke was in onwettige knoeiery en korrupsie. |
| Whisper Large v3 | Een paar weke gelede, nadat die informatie dier die journalist Makis Triantafilopoulos in sy gewilde televisieprogram Zungla op Alfa TV gepubliseer is, is een lid van die parlement en procureur Petros Mantovilos uit sy positie ontheef, aangesien lede van sy kabinet betrokke was in onwetige knoeireie en korruptie. |
Language detection
All 10 clips we sent without a language field were identified as Afrikaans. Auto-detection is safe here; sending language=af still skips the detection step.
Quickstart
The request is the OpenAI transcription request with a different base URL. language is optional.
curl https://api.inferenceapis.com/v1/audio/transcriptions \
-H "Authorization: Bearer $INFERENCE_API_KEY" \
-F "model=openai/whisper-large-v3" \
-F "language=af" \
-F "file=@recording.mp3" \
-F "response_format=verbose_json"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="af", 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 Afrikaans test.
FAQ
How accurate is Whisper for Afrikaans?
On 30 clean read-speech clips it made 30% word errors, about one word in 3. Gets the content across, but needs a human pass before it is published or quoted. Real recordings with noise, accents or several speakers will score worse.
Does language auto-detection work for Afrikaans?
In our test, yes: all 10 clips sent without a language were identified as Afrikaans. Sending language=af still saves the detection step on every request.
Can it translate Afrikaans speech to English?
The API returns Afrikaans text. For English, transcribe first and translate the text with a chat model on the same key.