← All examples

ModelEthio-ASR Amharic (600M)8Ethio-ASR Multi (300M)17Ethio-ASR Multi (600M)10Ethio-ASR Multi (94M)12Meta MMS-1B13OmniASR-CTC (1B)2OmniASR-CTC (300M)5OmniASR-CTC (3B)0OmniASR-LLM (1B)0OmniASR-LLM (300M)2OmniASR-LLM (3B)2OpenAI Whisper-small13
Audio qualityClean Studio5Ambient Room Noise5Harsh Environment6Distant / Low-Volume Mic4Wideband Phone (VoLTE)5Randomized Phone Channel5Narrowband 2G + Packet Loss10

What did the model get wrong?

Ethio-ASR Multi (600M) Narrowband 2G + Packet Loss   Example 19 of 20  Âˇ  speaker: Male

🎧 Listen — Narrowband 2G + Packet Loss
Heavy background noise plus phone-call compression — worst-case stress test.
✅ What was actually said
ቁጥሩ ሰባት ሶስት አንድ ዜሮ አንድ ሁለት ሰባት ዘጠኝ አምስት ዜሮ አራት ነው
🤖 What this model heard
amh ቁጡ ሰባት ቾ ን ዜሮ አንድ ሁለት ሰ ዘ ም ዜሮአራትራነ
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
4
Correct
7
Wrong
2
Missed
1
Extra
31%
Words right
Out of 13 spoken words, this model got 4 right (31%). Errors: 7 words wrong; 2 missed; 1 extra.

Try switching Audio quality — see the same model degrade as the audio gets noisier.