Inference APIs

Measured accuracy

ModelWord error rateCharacter error rateClips with no errorsEmpty transcriptsSpeed
Whisper Large v3
openai/whisper-large-v3
over 100%over 100%0 of 3029 of 300.165× real time

Returns nothing. The model lists this language, but our test clips came back as empty transcripts. 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.

Spokenአንድ ሰው በደቡብ አፍሪካ ውስጥ ላሉት ፓርኮችም ሆነ ለሁሉም የደቡብ አፍሪካ ብሔራዊ ፓርኮች መግቢያ ምርጫ የሚሰጥ የዱር ካርድ መግዛት ጠቃሚ ሊሆን ይችላል።
Whisper Large v3(empty transcript)

Language detection

All 10 clips we sent without a language field were identified as Amharic. Auto-detection is safe here; sending language=am still skips the detection step.

What to do instead

We would rather you know this before you build than after. For Amharic, look at services that train specifically on the language, such as Google Cloud Speech-to-Text or Azure AI Speech, both of which publish their language lists, or a Whisper checkpoint fine-tuned on Amharic from Hugging Face run on your own hardware. If your audio mixes Amharic with a well-supported language, the well-supported parts will still transcribe. You can confirm all of this on your own audio with the free credit on a new account: transcription playground.

FAQ

How accurate is Whisper for Amharic?

On this endpoint it is not usable: 29 of 30 Amharic test clips came back as empty transcripts.

Does language auto-detection work for Amharic?

In our test, yes: all 10 clips sent without a language were identified as Amharic. Sending language=am still saves the detection step on every request.

Can it translate Amharic speech to English?

The API returns Amharic text. For English, transcribe first and translate the text with a chat model on the same key; with accuracy this low the translation will inherit the errors.

Related languages: Somali (93%) · Swahili (34%)