Hemory listens, always on - wiring real memory into your agents
Hemory records and transcribes your everyday conversations into a private, searchable memory that connects to AI agents via MCP.
8
Views on SecondWing
1
Total wings
Read 6 times by AI crawlers since Sep 2026
Google · Anthropic
Hemory is an always-on listening app that turns real-world conversations into a private, searchable memory connected to AI agents over MCP.
Hemory listens to your conversations and turns them into transcripts and memories that you or your AI agents can search and build documents from.
Hemory comes from the words Hear and Memory. It is described as not meeting transcription, but a tool that captures every conversation and keeps it as memory, so you can revisit it and build on it through your AI agent.
Every listening session becomes an auto-transcribed, timestamped, speaker-tagged, AI-polished document. The day is automatically split into six activity types (meeting, work, meal, social, learning, chit-chat), each summarized according to its type. Hemory recognizes speakers by voice across sessions and includes a built-in Q&A agent for querying memories directly, as well as a standard MCP server so agents like Claude Code, Codex, Cursor and others can search and generate from the captured memories.
On privacy, the site states that audio is never stored in the cloud, is processed as a stream and destroyed the moment processing ends, and that raw audio is stored only on the listening device, never synced across devices. A self-host option is described as coming soon.
Audio is processed as a stream and destroyed immediately after processing, never stored in the cloud.
Raw audio stays local to the device that recorded it and is never synced across devices.
Listening hours are only consumed when someone is actually speaking; silence and background noise are excluded from the quota.
The site states user audio, transcripts, and voice feature data are never used to train or improve any AI model.
Unsubscribing only stops new transcription hours; previously generated memories and transcripts remain viewable until account deletion.
Users can optionally run Hemory entirely on their own infrastructure.
Free – $10/mo
Free plan available
iOS, macOS, Self-hosted, Windows, Linux, Web, Android
Every listening session becomes a timestamped, speaker-tagged, AI-polished transcript supporting 100 languages.
The day is automatically split into six recognized activity types on a timeline without manual selection.
Summaries follow the activity type, producing structured meeting minutes with agenda, decisions and action items, or light notes for meals and chit-chat.
Hemory recognizes voices across sessions; naming a speaker once updates all related entries, with confidence hints for unconfirmed speakers.
A multi-turn conversational agent lets users ask questions directly grounded in their transcripts.
A standard MCP server with a growing set of tools connects Hemory memories to mainstream AI agents.
GeminiMCP-connected agent client
CursorMCP-connected agent client$0one-time
Try it now, no payment.
$10/mo
The budget pick — just the conversations that matter.
$20/mo
For professional use, enough for everyday work.
Pricing last verified September 2026
View current pricingStructured information about this product, for people and AI systems.
Hemory is an always-on listening app that turns real-world conversations into a private, searchable memory connected to AI agents over MCP. Hemory listens to your conversations and turns them into transcripts and memories that you or your AI agents can search and build documents from.
How Hemory was built, and the people behind it.
0.0
Based on 0 reviews
Be the first to review Hemory.
Write a reviewMistral's multilingual AI assistant
xAI's conversational assistant with real-time X data
Prompt, run, edit, and deploy full-stack apps
Organize your work and life, finally
Build and deploy on the frontend cloud
