← All examples

ModelEthio-ASR Amharic (600M)4Ethio-ASR Multi (300M)5Ethio-ASR Multi (600M)4Ethio-ASR Multi (94M)4Meta MMS-1B9OmniASR-CTC (1B)5OmniASR-CTC (300M)7OmniASR-CTC (3B)0OmniASR-LLM (1B)0OmniASR-LLM (300M)0OmniASR-LLM (3B)0OpenAI Whisper-small12
Audio qualityClean Studio4Ambient Room Noise1Harsh Environment3Distant / Low-Volume Mic2Wideband Phone (VoLTE)2Randomized Phone Channel1Narrowband 2G + Packet Loss8

What did the model get wrong?

Ethio-ASR Amharic (600M) Clean Studio   Example 6 of 20  Β·  speaker: Male

🎧 Listen β€” Clean Studio
Clear, studio-quality recording. Best-case audio.
βœ… What was actually said
αˆ˜α‰Έ αŠα‹ α‹¨αˆšαˆ΅α‰°αŠ«αŠ¨αˆˆα‹ ሡራ αˆ΅αˆ‹αˆˆα‰₯ኝ α‰ΆαˆŽ α‰’αˆ†αŠ•αˆαŠ αŒ₯ሩ αŠα‹
πŸ€– What this model heard
αˆ˜α‰ΈαŠα‹ α‹¨αˆšαˆ΅α‰°αŠ«αŠ¨αˆˆα‹ ሡራ αˆ΅αˆ‹αˆˆα‰₯αŠ• α‰ΆαˆŽ α‰₯α‹¨αˆ†αŠ•αˆαŠ αŒ₯ሩ αŠα‹
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
5
Correct
3
Wrong
1
Missed
0
Extra
56%
Words right
Out of 9 spoken words, this model got 5 right (56%). Errors: 3 words wrong; 1 missed.

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