minutes

Comparison

Minutes vs MacWhisper

MacWhisper and Minutes are on the same side of the line that matters most: transcription runs locally on your Mac, and both can run Whisper or Parakeet models. The difference is the shape of the job. MacWhisper is the best drag-and-drop file transcriber on macOS — files in, transcripts and subtitles out. Minutes is a conversation memory layer — meetings and memos in, a growing structured archive out, one your AI agents can query. Respect where it's due; this is a comparison between two local-first tools, and many people legitimately want the other one.

Last reviewed: 2026-07-11Fit-based comparison

Quick verdict

Choose MacWhisper if your job is transcribing files — interviews, podcasts, videos, YouTube links — and you want the most polished Mac GUI for it, with subtitle export and a one-time price.

Choose Minutes if your job is remembering conversations — recording meetings and memos into a private, diarized, searchable archive that Claude and other agents can use — and you want it open source and free.

At A Glance
CategoryMacWhisperMinutes
Best forTranscribing audio/video files on a Mac with a polished GUIOn-device conversation memory: meetings, memos, and dictation your agents can query
Core jobFile in, transcript out — batch jobs, subtitles, YouTube and media-file URLsCapture conversations, diarize them, keep a searchable structured record
Where transcription runsOn-device (Whisper, Parakeet, and other local models)On-device (whisper.cpp or parakeet.cpp)
Optional cloud AIBYO API keys for summaries/chat (OpenAI, Anthropic, others) — or fully local via Ollama/LM StudioOptional and explicit: Claude via MCP, a local LLM (Ollama), or BYO-key cloud providers — off by default
Durable outputExports: txt, srt, vtt, md, pdf, docx — per fileMarkdown archive with YAML frontmatter, action items, decisions — per conversation, organized over time
Speaker handlingAutomatic speaker recognition (Pro)Diarization plus confidence-aware attribution that learns real names
Agent / MCP surfaceCLI control and workflow automations; no MCP server we could findMCP server (31 tools), CLI, SDK, Claude Code plugin over your local files
Open sourceNoYes, MIT
PlatformsmacOS (14+ for the App Store build) and iOS; no Windows or LinuxmacOS menu bar app + CLI (open source, builds from source elsewhere)
PricingFree tier; Pro €64 one-time direct (App Store channel sells subscriptions plus a pricier one-time lifetime unlock)Open source and free to run yourself

Where MacWhisper wins

  • File-transcription ergonomics are unmatched: drag-and-drop batches, YouTube and media-file URLs, podcast transcription with per-speaker files, filler-word removal, and a real subtitle workflow (.srt/.vtt with inline video preview and auto-translation).
  • One-time pricing (€64 direct, lifetime updates) is a genuinely fair deal, and the free tier already covers basic recording and file transcription in 100 languages.
  • An iOS companion app exists; Minutes is desktop-first today.

Where Minutes wins

  • It builds an archive, not just outputs: every meeting, memo, and dictation becomes structured markdown — attendees, action items, decisions in YAML — organized and greppable over months, not a folder of one-off exports.
  • It's agent-native: 31 MCP tools, a CLI, an SDK, and a Claude Code plugin let your assistant search meetings, track commitments, and build person profiles from your local files. MacWhisper automates workflows; it doesn't give agents a memory.
  • It's open source (MIT) and free — the entire pipeline is auditable Rust, which matters if 'local' is a compliance requirement rather than a preference.
Workflows

The overlap is real: both record meetings, both transcribe locally, both can use Whisper or Parakeet engines. The divergence is what happens after transcription. MacWhisper's output is a document you export and move somewhere; its center of gravity is the file. Minutes' output is an entry in a corpus — ~/meetings accumulates, search spans months, and MCP tools answer questions like 'what did we decide about pricing in April' across everything.

A fair test: open your transcription tool's output folder. If it's a pile of exports you rarely revisit, either tool works and MacWhisper is more polished. If you wish that pile were a queryable memory, that wish is the entire reason Minutes exists.

Which Should You Pick?

Pick MacWhisper for file work: interviews, podcast episodes, subtitle jobs, transcribing someone else's videos. It is the best tool on macOS for exactly that, and the one-time price is honest.

Pick Minutes for conversation memory: your own meetings and ideas, captured continuously into structured local files your agents read straight from disk — no vendor sync, no account.

Plenty of people should own both — they're neighbors, not rivals: one optimizes the transcript, the other optimizes the archive.

When Minutes Is Not The Right Fit

Minutes is not the right first choice if your work is transcribing files you receive — it's built around capturing live conversations, not batch-processing media libraries or producing subtitles.

It's also not the fit if you want an iOS-first experience or a polished GUI for one-off transcription jobs; MacWhisper is simply better at those today.

How We Evaluated

This is a fit-based comparison between two local-first tools, reviewed on 2026-07-11 against MacWhisper's official site and App Store listing, linked below. MacWhisper's local-by-default transcription, Whisper/Parakeet engine support, Pro feature list, and €64 one-time direct pricing (with a separate App Store channel selling subscriptions and a one-time lifetime unlock) are drawn from its own pages.

The Minutes side is grounded in its public docs and open-source repository. Both tools' privacy claims are architecture-level and, in Minutes' case, verifiable in source.

Next step

Sources