Czech speech-to-text
GoodWe sent 30 Czech test clips through this API. Whisper Large v3 made 12% word errors, about one word in 9; Parakeet made 14%. $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 | 12% | 3.1% | 7 of 30 | 0 of 30 | 0.082× real time |
Parakeet TDT 0.6B v3nvidia/parakeet-tdt-0.6b-v3 | 14% | 4.4% | 7 of 30 | 0 of 30 | 0.047× 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 | Termín safari ve známém slova smyslu označuje pozemní cestování za účelem prohlídky ohromující africké zvěře, obzvláště na savaně. |
| Whisper Large v3 | Termín safari ve známém slova smyslu označuje pozemní cestování za účelem prohlídky ohromojící africké zvěře, obzvláště na savani. |
| Parakeet (its median clip) | Spoken: Poté byl přesunut do Addenbrookovy nemocnice v Cambridge. Poté byl přesunut do Edinburghovy nemocnice v Cambridge. |
Language detection
Send language=cs. Of 10 clips sent without it, only 8 were identified as Czech; the others were labelled Slovak (1), English (1). When detection is wrong, Whisper writes the transcript in the language it guessed.
Which model for Czech
Whisper Large v3 was more accurate on these clips (12% against 14%). Parakeet is faster, which matters mainly on long recordings. Both cost the same. Side by side.
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 Czech.
curl https://api.inferenceapis.com/v1/audio/transcriptions \
-H "Authorization: Bearer $INFERENCE_API_KEY" \
-F "model=openai/whisper-large-v3" \
-F "language=cs" \
-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="cs", 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 Czech test.
FAQ
How accurate is Whisper for Czech?
On 30 clean read-speech clips it made 12% word errors, about one word in 9. 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 Czech?
Not reliably. Of 10 clips sent without a language, 8 were identified as Czech; the rest were labelled Slovak (1), English (1). A wrong label means the transcript is written in the wrong language, so always send language=cs.
Whisper or Parakeet for Czech?
On the same clips Whisper Large v3 scored 12% and Parakeet 14%. Whisper was more accurate here; Parakeet is faster, which matters mainly for long recordings. Both cost $0.002 per minute.
Can it translate Czech speech to English?
The API returns Czech text. For English, transcribe first and translate the text with a chat model on the same key.