โ† All examples

ModelEthio-ASR Amharic (600M)1Ethio-ASR Multi (300M)7Ethio-ASR Multi (600M)3Ethio-ASR Multi (94M)7Meta MMS-1B9OmniASR-CTC (1B)3OmniASR-CTC (300M)5OmniASR-CTC (3B)4OmniASR-LLM (1B)2OmniASR-LLM (300M)4OmniASR-LLM (3B)2OpenAI Whisper-small9
Audio qualityClean Studio4Ambient Room Noise2Harsh Environment4Distant / Low-Volume Mic4Wideband Phone (VoLTE)3Randomized Phone Channel3Narrowband 2G + Packet Loss9

What did the model get wrong?

OmniASR-CTC (1B) Wideband Phone (VoLTE)   Example 7 of 20  ยท  speaker: Female

๐ŸŽง Listen โ€” Wideband Phone (VoLTE)
Compressed, narrow-band audio like a regular phone call.
โœ… What was actually said
แ‹จแŠขแŠ•แ‰ฐแˆญแŠ”แ‰ต แŒฅแ‰…แˆ แ‹ˆแ‹ญแˆ แ“แŠฌแŒ… แˆ˜แŒแ‹›แ‰ต แŠฅแ‰ฝแˆ‹แˆˆแˆ แ‹‹แŒ‹แ‹ แˆตแŠ•แ‰ต แŠแ‹
๐Ÿค– What this model heard
แ‹จ แŠขแŠ•แ‰ฐแˆญแŠ”แ‰ต แŒฅแ‰…แˆ แ‹ˆแ‹ญแˆ แ“แŠฌแŒ… แˆ˜แŒแ‹›แ‰ต แŠฅแ‰ฝแˆ‹แˆˆแŠ• แ‹‹แŒ‹แ‹ แˆตแŠ•แ‰ต แŠแ‹
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
7
Correct
2
Wrong
0
Missed
1
Extra
78%
Words right
Out of 9 spoken words, this model got 7 right (78%). Errors: 2 words wrong; 1 extra.

Try switching Audio quality โ€” see the same model degrade as the audio gets noisier.