Take audit: 8T9yec9JZ-y1qKPALIXT7
2026-08-22T09:22:23.810Z — ray
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 · 630 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_auto_bare_deg gpt-4o-mini-transcribe · degraded · auto · bare
— MER accuracy pending Ground Truth · 827 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_auto_bare_clean gpt-4o-mini-transcribe · clean · auto · bare
— MER accuracy pending Ground Truth · 579 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_auto_prompt_deg gpt-4o-mini-transcribe · degraded · auto · mixed prompt
— MER accuracy pending Ground Truth · 707 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_auto_prompt_clean gpt-4o-mini-transcribe · clean · auto · mixed prompt
— MER accuracy pending Ground Truth · 769 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_auto_prompt_64k gpt-4o-mini-transcribe · 64k · auto · mixed prompt
— MER accuracy pending Ground Truth · 715 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_zh_prompt_deg gpt-4o-mini-transcribe · degraded · zh · mixed prompt
— MER accuracy pending Ground Truth · 627 ms · verdict: unjudged
老師,I want a juice, too, please.
mini_zh_prompt_clean gpt-4o-mini-transcribe · clean · zh · mixed prompt
— MER accuracy pending Ground Truth · 920 ms · verdict: unjudged
老師,I want a juice, too, please.
gpt_tw_en_bare_deg gpt-transcribe · degraded · zh-tw + en · bare
— MER accuracy pending Ground Truth · 618 ms · verdict: unjudged
老師,I want juice, two please.
gpt_tw_en_prompt_deg gpt-transcribe · degraded · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 857 ms · verdict: unjudged
老師,I want juice too, please.
gpt_tw_en_prompt_clean gpt-transcribe · clean · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 582 ms · verdict: unjudged
老師,I want juice, two, please.
gpt_tw_en_prompt_64k gpt-transcribe · 64k · zh-tw + en · mixed prompt
— MER accuracy pending Ground Truth · 570 ms · verdict: unjudged
老師,I want the juice too, please.
gpt_tw_only_prompt_deg gpt-transcribe · degraded · zh-tw · mixed prompt
— MER accuracy pending Ground Truth · 708 ms · verdict: unjudged
老師,I want juice too, please.
Winning configuration
Record the variant and rationale to hand to qa-rails after reviewing this Take.