← All examples

ModelEthio-ASR Amharic (600M)5Ethio-ASR Multi (300M)10Ethio-ASR Multi (600M)11Ethio-ASR Multi (94M)10Meta MMS-1B12OmniASR-CTC (1B)6OmniASR-CTC (300M)5OmniASR-CTC (3B)1OmniASR-LLM (1B)2OmniASR-LLM (300M)7OmniASR-LLM (3B)6OpenAI Whisper-small12
Audio qualityClean Studio4Ambient Room Noise4Harsh Environment3Distant / Low-Volume Mic4Wideband Phone (VoLTE)3Randomized Phone Channel2Narrowband 2G + Packet Loss10

What did the model get wrong?

Ethio-ASR Multi (94M) Narrowband 2G + Packet Loss   Example 5 of 20  ·  speaker: Male

🎧 Listen — Narrowband 2G + Packet Loss
Heavy background noise plus phone-call compression — worst-case stress test.
✅ What was actually said
ቁጥሬ ዜሮ ዘጠኝ ሰላሳ አራት አራት አምስት ስድስት ሰባት ስምንት ዘጠኝ ነው
🤖 What this model heard
amh ቁጥረ በበጣኝ ሰላሰአት አራት አምስ ድርሰባት ምንት በጣኝ ነው
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
2
Correct
8
Wrong
2
Missed
0
Extra
17%
Words right
Out of 12 spoken words, this model got 2 right (17%). Errors: 8 words wrong; 2 missed.

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