Take audit: 9KNliQqS2eFeD_CeQPb7G
2026-08-22T09:05:27.536Z — Plain
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 · 1262 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_auto_bare_deg gpt-4o-mini-transcribe · degraded · auto · bare
— MER accuracy pending Ground Truth · 1349 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_auto_bare_clean gpt-4o-mini-transcribe · clean · auto · bare
— MER accuracy pending Ground Truth · 1570 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_auto_prompt_deg gpt-4o-mini-transcribe · degraded · auto · mixed prompt
— MER accuracy pending Ground Truth · 1350 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_auto_prompt_clean gpt-4o-mini-transcribe · clean · auto · mixed prompt
— MER accuracy pending Ground Truth · 1560 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_auto_prompt_64k gpt-4o-mini-transcribe · 64k · auto · mixed prompt
— MER accuracy pending Ground Truth · 2481 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_zh_prompt_deg gpt-4o-mini-transcribe · degraded · zh · mixed prompt
— MER accuracy pending Ground Truth · 1378 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
mini_zh_prompt_clean gpt-4o-mini-transcribe · clean · zh · mixed prompt
— MER accuracy pending Ground Truth · 1419 ms · verdict: unjudged
我喜歡喝juice,很好喝喔
gpt_tw_en_bare_deg gpt-transcribe · degraded · zh-tw + en · bare
— MER accuracy pending Ground Truth · 2296 ms · verdict: unjudged
我喜歡喝 juice,很好喝喔。
gpt_tw_en_prompt_deg gpt-transcribe · degraded · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 1521 ms · verdict: unjudged
我喜歡喝 juice,很好喝喔。
gpt_tw_en_prompt_clean gpt-transcribe · clean · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 1874 ms · verdict: unjudged
我喜歡喝 juice,很好喝喔。
gpt_tw_en_prompt_64k gpt-transcribe · 64k · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 1624 ms · verdict: unjudged
我喜歡喝 juice,很好喝喔。
gpt_tw_only_prompt_deg gpt-transcribe · degraded · zh-tw · mixed prompt
— MER accuracy pending Ground Truth · 1506 ms · verdict: unjudged
我喜歡喝 juice,很好喝喔。
Winning configuration
Record the variant and rationale to hand to qa-rails after reviewing this Take.