raw Nepali audio infrastructure — speech in, speech out
research Help us measure how small speech models actually sound. About five minutes.
| clips | 20 short samples, drawn from 12 sentences, in random order |
| your job | rate each one from 1 (bad) to 5 (excellent) on how natural and clear it sounds |
| please | use headphones somewhere quiet, and judge the audio quality — not whether you like the voice |
© 2026 ampixa labs.
We store only the display name you type, your ratings and their timestamps.
No email, no account, no analytics, no IP logging.