giw-gen5 console

run history staging

Run 21159ad5

$0.000717
total cost
6,173
input tokens
41%
cached
2
turns
active
status

arc peak · pack behavioral.personal@1 · voice default · stage early · tier free · mode assessor

Total cost

componentshareusd
LLM measured billed by the provider, reconcilable against the gateway log 10.4%$0.000717
TTS estimated 221 chars the interviewer would speak 48.2%$0.003315
STT estimated 22s of candidate audio · Cloudflare Workers AI — @cf/openai/whisper 41.4%$0.002846
total $0.006877

TTS $15/1M chars · STT $0.00762/audio-min · speech figures are ESTIMATES from the transcript — no audio was produced or transcribed

STT rate source: neuron rate inherited from probe-atlas; neurons-per-minute is an unverified guess

Transcript & scoring

turn 1 · pull · motivation · core

Q. What's drawing you to this role?

That's a really clear way to put it—liking that the failure was concrete.

A. I moved into backend work after a data migration went badly on my second project. We lost two days because nobody had checked the encoding on the source files. I ended up owning the import path after that, and I liked that the failure was concrete - you could see exactly which row broke and why.

score 5.5/10 solid · llm $0.000562 (3797 in, 256 cached, 235 out) · tts 105ch $0.001575 · stt 22s $0.002846
The answer gives a concrete, personal story with a real reason for backend interest, but it lacks any tie to this specific company or role, and the motivation is backward-looking rather than forward-looking.
turn 2 · evidence · self_awareness · core

Q. What's an area you're actively trying to improve in your work?

A. (unanswered)

· tts 62ch $0.000930 · stt 0s $0.000000

Cost by stage

stagecallswasted incachedthink outcost
analyze_turn1 19380 0182 $0.000322
generate_questions1 23762304 0492 $0.000154
interviewer_turn1 1859256 053 $0.000240