Help

Fix it in a minute

Nearly everything people write in about is one of the problems below, and all of them have a fix you can apply right now. If yours is not here, write to us — a real person reads it.

First run

Getting started

AI features are locked and I have already signed up

Confirming your email address is what unlocks AI. It is how we stop automated signups draining the free allowance, which is what lets the free tier exist at all.

01Check your inbox for the confirmation email, including spam and Promotions.
02Click the link. The desktop app picks up the change within a few seconds.
03No email after a few minutes? Sign in on the website and request a new one from the banner at the top.

The app cannot hear me

macOS holds microphone access per app, and a denied prompt is silent afterwards — nothing tells you it was refused.

01Open System Settings → Privacy & Security → Microphone.
02Turn Meeting Flow on. If it is not listed, quit and reopen the app so it asks again.
03Check Settings → Audio Sources in the app: the Microphone row should read Active while you talk.

It transcribes me but not the other person

The other party's voice comes from system audio, which needs screen-recording permission on macOS — a separate grant from the microphone.

01Open System Settings → Privacy & Security → Screen & System Audio Recording.
02Enable Meeting Flow, then quit and reopen the app — macOS only applies this on relaunch.
03In Settings → Audio Sources, switch System Audio on and confirm the pill reads Active.

During a meeting

Transcription and answers

The transcript stops partway through a meeting

The live transcription socket is renewed periodically, and a network drop during a renewal falls back to the slower batch engine rather than stopping.

01Check your connection — a VPN or a captive Wi-Fi portal is the usual culprit.
02Look for the warning in the app: it says explicitly when it has fallen back.
03Stop and restart listening. If it recurs on the same network, switch to a local engine in Settings → Speech Recognition.

Answers take too long to appear

Model choice and transcription engine both affect this, and the batch engines add a fixed wait because they cannot tell when you have finished speaking.

01Settings → Speech Recognition: choose Deepgram Flux. It detects end-of-turn itself and removes the largest fixed delay.
02Settings → AI Model: prefer a fast model. Pro-tier reasoning models are slower by design.
03Shorten your context profile. Very long résumé and job-description text adds latency to every single answer.

Suggested follow-ups stopped appearing mid-call

This is deliberate. When an account is sending requests unusually fast, background features pause first so the answer you are waiting on stays fast.

01Nothing to do — they resume on their own within a minute.
02If it happens constantly, something is retriggering requests. Tell us what you were doing and we will look.

The wrong person is credited with what was said

Speaker attribution compares the loudness of your microphone against the incoming audio, so headphones-off speaker playback can bleed into your mic.

01Use headphones. This resolves nearly every case.
02Move the microphone away from your speakers, or lower the call volume.

Plans

Credits and billing

My credits ran out faster than I expected

Live transcription is the largest cost by a wide margin — far more than the AI answers. Two audio lanes cost roughly twice one.

01Run one lane when you only need the other person transcribed.
02Switch to the on-device engine in Settings → Speech Recognition. It costs no credits at all.
03Check Dashboard → Usage to see exactly where the month went.

See your usage

I need more credits before my cycle resets

Top-up packs exist for this. Pack credits never expire and are used only after the monthly allowance is gone.

01Open Dashboard → Billing and choose a pack.
02If you are regularly running out, upgrading is better value per credit than repeat top-ups.

Compare plans

Screen capture says I need to upgrade

Screen capture and multimodal analysis are available from Starter upward. Pro and Premium additionally get the sharper vision model.

01Check which plan you are on in Dashboard → Billing.
02Starter and above unlock screen capture; Pro and above unlock advanced screen and code analysis.

Compare plans

I want to cancel

Cancellation is self-service and takes effect at the end of the period you have already paid for.

01Open Dashboard → Billing and use the Stripe customer portal.
02Access continues until the current period ends; nothing renews after that.

Refund policy

Privacy

Audio and data

I do not want my audio leaving my computer

The default engine is a cloud service, because it is the only one fast enough to answer the moment a question ends. There is a fully local alternative.

01Open Settings → Speech Recognition.
02Choose Local Whisper Model or Apple System Dictation. Both are labelled as staying on your Mac.
03Answers still require sending transcript text to a model provider — only audio stays local.

How we handle audio

I want everything deleted

Account deletion is immediate and irreversible, and removes every record tied to your user id.

01Open Dashboard → Account and use the Danger zone.
02Stripe keeps invoice records separately, as tax law requires.

Still stuck

Send us your operating system, the app version from Settings → About, and what you were doing when it happened. That is usually enough to answer on the first reply instead of the third.

Email support