Take audit: hoMHBcPb7kPMQy24pQ4GC
2026-08-22T09:43:31.771Z — 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 · 588 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_auto_bare_deg gpt-4o-mini-transcribe · degraded · auto · bare
— MER accuracy pending Ground Truth · 479 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_auto_bare_clean gpt-4o-mini-transcribe · clean · auto · bare
— MER accuracy pending Ground Truth · 735 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_auto_prompt_deg gpt-4o-mini-transcribe · degraded · auto · mixed prompt
— MER accuracy pending Ground Truth · 735 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_auto_prompt_clean gpt-4o-mini-transcribe · clean · auto · mixed prompt
— MER accuracy pending Ground Truth · 763 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_auto_prompt_64k gpt-4o-mini-transcribe · 64k · auto · mixed prompt
— MER accuracy pending Ground Truth · 819 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_zh_prompt_deg gpt-4o-mini-transcribe · degraded · zh · mixed prompt
— MER accuracy pending Ground Truth · 706 ms · verdict: unjudged
老師,可以請你再說一次嗎?
mini_zh_prompt_clean gpt-4o-mini-transcribe · clean · zh · mixed prompt
— MER accuracy pending Ground Truth · 907 ms · verdict: unjudged
老師,可以請你再說一次嗎?
gpt_tw_en_bare_deg gpt-transcribe · degraded · zh-tw + en · bare
— MER accuracy pending Ground Truth · 703 ms · verdict: unjudged
老師可以請你再說一次嗎？
gpt_tw_en_prompt_deg gpt-transcribe · degraded · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 622 ms · verdict: unjudged
老師可以請你再說一次嗎？
gpt_tw_en_prompt_clean gpt-transcribe · clean · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 788 ms · verdict: unjudged
老師可以請你再說一次嗎？
gpt_tw_en_prompt_64k gpt-transcribe · 64k · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 726 ms · verdict: unjudged
老師可以請你再說一次嗎？
gpt_tw_only_prompt_deg gpt-transcribe · degraded · zh-tw · mixed prompt
— MER accuracy pending Ground Truth · 934 ms · verdict: unjudged
老師可以請你再說一次嗎？
Winning configuration
Record the variant and rationale to hand to qa-rails after reviewing this Take.