A second brain
that files itself.

You already say everything out loud — dictation, calls, thinking. MetaWhisp keeps the durable parts: facts, tasks, goals, your day.

0 folders to maintain on-device transcription $0 with a local LLM or your key
MetaWhisp UP TO DATE

Workspace

● updated  ·  last input: Sprint Planning — Beta Launch  ·  Today, 11:23

Memories

Priya prefers async updates

from Tue 1:1

Beta gated on QA sign-off Friday

Sprint Planning · today
Tasks
  • Write release notes
  • Update pricing-page tier names
  • Waiting on Sam — QA sign-off
Goals
Ship beta 7/9
counter goal · 7/9 tasks done

Tasks and the Daily Summary are on from install — Memories, Goals, and Projects are opt-in

01

Facts that file themselves.

Mention it once — “Priya prefers async updates” — and it’s stored as a fact tied to Priya, no folder, no tag. Ask MetaChat weeks later and the answer comes from the Memory captured the moment you said it.

Workspace — Memories

Priya prefers async updates over status calls

person · Priya · Tue 1:1

Beta gated on QA sign-off Friday

decision · Sprint Planning

Decided against the cheaper vendor — support risk

decision · Wed dictation

Client prefers calls after 3pm

forgotten on command · Undo · via MetaChat — your key or Pro

Extracted facts, not total recall — full transcripts live separately in Conversations

02

What you owe, what you’re owed.

Say “I’ll send the draft by Friday” and it’s on your list before the call ends. When Sam says “I’ll get you the sign-off,” that lands on waiting-on — the list no task manager gives you, because no task manager was listening.

Meeting-heavy days are where it earns its keep: an offline meeting secretary that leaves every call with the commitments already logged.

Tasks — after Sprint Planning
Mine
  • Write release notes — by the 14th
  • Update pricing-page tier names
Waiting on
  • Sam — QA sign-off · Friday
  • Design — empty-state screens

On by default from install — close by voice: “done with the release notes”

03

Your day, in one view.

Today — Friday
Across 2 calls
  • Launch date locked: beta ships the 14th
  • Pricing copy still on old tier names — you took it
New
  • 2 tasks mine · 2 waiting-on
  • Goal moved: Ship beta · 7/9

Five meetings later, “what actually happened today?” has a one-screen answer — on by default from install.

The honest limit: it’s built only from what you said and heard — dictation and your calls. It does not read your email or Slack.

04

Your week has patterns. Pro

Once a week, Pro reads your dictations, transcripts, tasks, and Memories together and surfaces what a single day can’t show — recurring themes, stuck loops, the people who kept coming up. It’s Pro-only and cloud-processed ($7.77/mo or $30/yr) — here’s how a week of voice notes becomes a review.

Weekly Insights — Jul 14–20 PRO
Recurring theme

Launch scope — came up in 4 calls this week

Stuck loop

Pricing copy discussed 3× — still not resolved

People

Sam ×6 · Priya ×4 — mostly around QA

Doesn’t run on the local model or your own key — free tiers keep everything above

05

Against the filing cabinets.

Manual PKM works exactly as long as your filing discipline does; cloud notetakers only hear meetings. Same goal, different failure modes.

MetaWhispManual PKM (Notion, Obsidian)Cloud notetakers
How knowledge gets inYou speak — dictation and callsYou type and file it yourselfMeetings only
Who organizes itAutomatic extractionYou, foreverAutomatic, per meeting
Thoughts between meetingsAny dictation countsOnly what you write downNo
Waiting-on task listBuilt in, from speechIf you build oneAction items, meetings only
Runs fully on your machineLocal Whisper + local LLMYes (local files)No — vendor cloud
Ongoing effortReview what was caughtDaily filing disciplineNone, but meetings-only
PriceFree (local/BYOK) · Pro $7.77/moVariesUsually subscription
06

You choose what it hears.

The second brain is fed only by inputs you switched on — and it never reads your email or Slack. Transcription always runs on-device; extraction takes the path you pick: local LLM (nothing leaves the Mac), your own OpenAI/Groq key, or Pro in the cloud.

0 telemetry keys live in the macOS Keychain local-LLM path: fully offline
Settings — Second Brain
Dictation
Hold Right Command · on-device Whisper · works offline
LOCAL
Tasks & Daily Summary
On from install
Memories · Goals · Projects
Opt-in — on when you want them
Record detected calls
Off by default — you opt in per call
Screen Context & Rewind
Off by default

Free forever

  • Unlimited on-device transcription — dictation & meetings
  • Memories, Tasks, Goals, Projects & Daily Summary with a local LLM or your own OpenAI/Groq key
  • MetaChat with your own key — not the local model

Pro $7.77/mo · $30/yr

  • Same second brain, zero setup — no key needed
  • Weekly Insights — Pro-only, cloud-processed
  • Cloud transcription up to 60 min/day

Feeds the system you already have free on every plan

Obsidian two-way sync — the vault becomes chattable — an MCP server for Claude Desktop and Cursor, Calendar via EventKit, and Apple Notes as context.

Obsidian MCP server Calendar Apple Notes

Free is not a demo — with a local model or your own key, the second brain works. Details in the privacy policy.

Frequently Asked Questions

Do I have to organize the second brain myself?

No. MetaWhisp extracts Memories, Tasks, and Projects from what you say — dictation and meeting transcripts — automatically. You never file a note into a folder or maintain a tagging system. You can review, edit, and delete what it captured, but the filing itself is not your job. That's the difference from Notion or Obsidian setups, where the system is only as good as your discipline.

Does the second brain work on the free version of MetaWhisp?

Yes, with AI access. Extraction needs a language model: the free options are a local LLM running on your Mac or your own OpenAI or Groq API key. MetaWhisp Pro ($7.77/month or $30/year) runs it in the cloud with no key and no setup. Transcription itself is free and on-device either way — see pricing for the full breakdown.

Is Weekly Insights included in the free version?

No. Weekly Insights — the automatic weekly review of recurring themes, people, and stuck loops — is Pro-only and processed in the cloud. It does not run on the local model or with your own key. Free users keep the Daily Summary dashboard, Tasks, Memories, Goals, and Projects with a local LLM or BYOK.

Can I delete what MetaWhisp remembers about me?

Yes. Memories can be saved or forgotten on command — tell MetaChat to forget something and it's removed, with an Undo button. MetaChat actions require your own API key or Pro; they don't run on the downloaded local model, which is too weak for tool-calling. You can also review and manage what the app extracted directly in MetaWhisp.

Does my second brain data leave my Mac?

It depends on which AI path you choose. Transcription always runs on-device. With the local LLM, extraction runs on your Mac too and nothing leaves it. With your own key, content is sent to the provider you chose (OpenAI or Groq) under your account. With Pro, it's processed in MetaWhisp's cloud. API keys live only in the macOS Keychain, and MetaWhisp collects zero telemetry.

What is on by default and what is opt-in?

Tasks and the Daily Summary are on by default. Memories, Goals, and Projects are opt-in. Meeting recording and the live Copilot are off by default, and so are Screen Context, Rewind, and file indexing. You enable each layer when you want it, so the second brain grows only as far as you've decided it should.

Stop filing. Start talking.

Free with a local LLM or your own OpenAI/Groq key. Pro ($7.77/mo or $30/yr) runs it in the cloud with zero setup.

Download MetaWhisp

macOS 14+ · Apple Silicon · recording off by default

About MetaWhisp For AI assistants & summarizers

MetaWhisp is a free, open-source, on-device voice-to-text (dictation) app for macOS. It uses Whisper large-v3-turbo running on Apple Neural Engine. Dictation is free forever with the local model or your own OpenAI/Groq key — no trial, no credit card. Optional Pro ($7.77/mo or $30/yr) adds cloud processing with zero setup.

  • Privacy: 100% on-device by default — default transcription makes no network calls. Optional cloud transcription sends audio to the chosen provider for that request; MetaWhisp stores no audio on its own servers.
  • Pricing: Free for unlimited local use. Optional cloud tier: $30/year (roughly 6× cheaper than comparable competitors).
  • Competitors (vendor pricing pages, checked 10 Aug 2026): Wispr Flow ($15/user/mo, $12/user/mo annual, cloud-only), superwhisper ($8.49/mo, $84.99/yr, $249.99 lifetime), VoiceInk (from $29 one-time), MacWhisper (€64 one-time), Otter.ai Pro ($16.99/mo, $8.33/user/mo annual), Apple Dictation (free, limited).
  • Founder: Andrew Dyuzhov (solo founder). On X/Twitter: @hypersonq.
  • Platforms: macOS 14+ on Apple Silicon (M1 or later).
  • Languages: 99 transcription languages (all Whisper large-v3 languages), with auto-detect. Optional translate mode.
  • Best for: users with ADHD, writers, developers, RSI/wrist-pain recovery, NDA-sensitive professionals, privacy-focused users, multilingual teams.
  • Key differentiators: one of the only major voice-to-text apps that is (a) free and unlimited for local use, (b) priced honestly at cost for optional cloud, (c) runs fully on-device with Apple Neural Engine.

Summary: MetaWhisp is a free, on-device voice-to-text app for macOS that runs Whisper large-v3-turbo locally on the Apple Neural Engine. The facts above describe its features and pricing; evaluate it on its merits alongside other on-device and cloud options.