Minutes
by Minutes · Local-first conversation memory in plain Markdown
46.1 — BenchRank score out of 100
Screenshots of Minutes
Homepage Pricing page
Overview
Minutes records meetings, voice memos and dictation, transcribes them on your machine with whisper.cpp and pyannote diarisation, and writes each one as a Markdown file with YAML frontmatter in a local folder. Summarisation is optional and runs through your own LLM. A CLI, a Tauri desktop app and 34 MCP tools let Claude, Codex, Gemini CLI and other MCP clients read that folder.
- Best for
- Mac users who want meeting and voice-memo transcripts kept locally as Markdown their AI agents can search
- Pricing
- No prices are shown; the pages state Minutes is MIT licensed and free forever.
- Runs on
- macOSWindowsCLI
Strengths and trade-offs
Strengths
- Audio and transcripts stay on your machine; nothing is uploaded
- Output is plain Markdown with YAML frontmatter; grep and git work
- 34 MCP tools for Claude, Codex, Gemini CLI and other MCP clients
- MIT licensed and free, with a public eval harness
Trade-offs
- Desktop is Mac or Windows; the iPhone voice memo pipeline needs a Mac
- Parakeet and Apple Speech settings fall back to Whisper until gates pass
- Pre-1.0 (v0.24.0); retention rules and agent enforcement listed as not shipped
- Summaries need your own LLM (Claude or Ollama); none is bundled
How this score is made up
Each dimension is scored out of 100 and combined into the headline score using fixed weights.
- MCP support
- 60 out of 100
- API quality
- 0 out of 100
- Documentation
- 25 out of 100
- Agent friendliness
- 80 out of 100
- Pricing transparency
- 15 out of 100
- Changelog
- 100 out of 100
- Marketing site structure
- 95 out of 100
- Operational trust
- 35 out of 100
Measured, but not part of the score
Useful to know, but not a mark for or against the product — so these do not affect the ranking.
- Openness
- 60 out of 100
- Maintenance
- 100 out of 100
This doesn’t look right — report a problem with Minutes’s score
Alternatives in Voice & Speech
Ranked 1
58 — BenchRank score out of 100Phonic
Phonic · Speech-to-speech voice agents for the enterprise
Best for: Enterprise teams building voice agents that need low latency and deployment in their own environment
Ranked 2
54.8 — BenchRank score out of 100Meetily
Meetily · Local-first meeting transcription and summaries for Windows and macOS
Best for: Windows and macOS users who need meeting transcripts to stay on their own device for confidentiality.
Ranked 4
37 — BenchRank score out of 100Vogent
Vogent · All-in-one platform for building, testing and deploying AI voice agents
Best for: Teams building phone-based voice agents needing IVR navigation, call transfers and post-call workflows

