Changelog
Every release since launch: kernel updates, detection-vector patches, and product changes, with date and scope. Proof we're still shipping.
Windows: answers start sooner. A short version you can say right away comes first, and the full answer picks up where it leaves off; behavioral questions get answers you can say in your own words; asking what to say always gets an answer; sharper screen reading, code tracing and live hints.
Windows: Aceloop is now an agent. One line in the chat header says what it is doing and what waits on your key; clicks on its buttons never reach the page; smoother listening and thinking; Solve works from a problem read off the screen; the window always fits its screen.
Windows: notes under each message (summary, math check, hint); short tips that teach the basics first; Run an example offers to Debug when it finds a bug.
Windows: help while you talk (hints, math checks, the parts of a question, what you promised to come back to, questions to ask them); Run an example traces your code step by step; every action rebindable in Settings, Keyboard; the right model for each job.
Windows: separate interview and reading languages. Answers come in the interview's language; you read everything in yours.
Windows: chat and Notes side by side, always; resume an interrupted interview and pick up where you left off.
Windows: a written-language setting, saved sessions that expire after 12 hours, and a launcher that keeps itself up to date.
Windows: window and text sizes that follow your screen.
Windows: 18 interview languages, right-to-left Hebrew, bold main points, and hover explanations for unfamiliar terms.
Windows: a new conversation engine, automatic model routing, the talk-track checklist, and system design that reads your whiteboard.
Windows: the Notes notebook beside the chat, the What I see page, and sharper speech recognition.
Windows: the chat-first layout.
Keep Error Lens cards horizontally stable while scrolling code.
Six-model Solve lineup, live usage meter, and streamlined Panda chat.
Chat UX overhaul: top-anchored reading with a reader-held scroll, collapsed-bubble recall, and a calmer attention model.
Voice smart answer + recap lanes upgraded to GPT-5.6 Terra (shadow-eval winner: better answers, ~2.4x faster, cheaper).
Q1 deep-arbiter precedence for updated findings + Q2 multi-line error-lens (spans, multi-row diff card, narrowest-wins overlap).
Error-lens: findings shown on the editor (severity squiggles + diff card at the flagged line, not just the rail). Live utterance chat ordering + agent-gated keep; reviewer truthfulness + dedupe hardening.
Live utterance zone: kept bubbles now follow the assist gate's real judgment (agent-officialized history); live-voice mint backoff + STT contextual priming.
Live utterance zone in chat (tentative + kept bubbles, real STT), per-interview-type transcription tuning, smoother live-voice reliability (mint backoff).
Live transcription indicator: a subtle activity trace by the Listening pill shows the words Aceloop is hearing in real time (interviewer and you in distinct colors).
Reliability and polish pass.