Take audit: QUmFEpq5DETCRP2gN5ihT
2026-08-22T09:32:56.986Z — Mika
0/13 judged · keyboard: Space play/pause, ↑/↓ focus, P pass, T Translation failure, F fail
Do a blind check (transcripts hidden)
Ground Truth
What the Tester actually said. Never pre-filled from an Engine — type what you hear.
Transcripts
mini_prod_baseline gpt-4o-mini-transcribe · degraded · en · bare
— MER accuracy pending Ground Truth · 624 ms · verdict: unjudged
This is for you, too.謝謝老師。
mini_auto_bare_deg gpt-4o-mini-transcribe · degraded · auto · bare
— MER accuracy pending Ground Truth · 664 ms · verdict: unjudged
This is for you, too.謝謝老師。
mini_auto_bare_clean gpt-4o-mini-transcribe · clean · auto · bare
— MER accuracy pending Ground Truth · 727 ms · verdict: unjudged
您S保佑,祝,謝謝老師。
mini_auto_prompt_deg gpt-4o-mini-transcribe · degraded · auto · mixed prompt
— MER accuracy pending Ground Truth · 652 ms · verdict: unjudged
This is for you too, 謝謝老師。
mini_auto_prompt_clean gpt-4o-mini-transcribe · clean · auto · mixed prompt
— MER accuracy pending Ground Truth · 715 ms · verdict: unjudged
Next is for you, 祝,謝謝老師。
mini_auto_prompt_64k gpt-4o-mini-transcribe · 64k · auto · mixed prompt
— MER accuracy pending Ground Truth · 655 ms · verdict: unjudged
This is for you, too. 謝謝老師
mini_zh_prompt_deg gpt-4o-mini-transcribe · degraded · zh · mixed prompt
— MER accuracy pending Ground Truth · 737 ms · verdict: unjudged
This is for you too. 謝謝老師。
mini_zh_prompt_clean gpt-4o-mini-transcribe · clean · zh · mixed prompt
— MER accuracy pending Ground Truth · 755 ms · verdict: unjudged
Next is for you, 謝謝老師。
gpt_tw_en_bare_deg gpt-transcribe · degraded · zh-tw + en · bare
— MER accuracy pending Ground Truth · 780 ms · verdict: unjudged
This is for you, too. 謝謝老師。
gpt_tw_en_prompt_deg gpt-transcribe · degraded · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 604 ms · verdict: unjudged
This is for you, too. 謝謝老師。
gpt_tw_en_prompt_clean gpt-transcribe · clean · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 840 ms · verdict: unjudged
This is for you, too. 謝謝老師。
gpt_tw_en_prompt_64k gpt-transcribe · 64k · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 1075 ms · verdict: unjudged
This is for you, too. 謝謝老師。
gpt_tw_only_prompt_deg gpt-transcribe · degraded · zh-tw · mixed prompt
— MER accuracy pending Ground Truth · 1036 ms · verdict: unjudged
This is for you, too. 謝謝老師。
Winning configuration
Record the variant and rationale to hand to qa-rails after reviewing this Take.