| What it does | A personal AI presence that remembers you, stays with you across devices, and asks before it acts
| A lightweight, secure personal AI agent you run on your own machine with your keys or local models
|
|---|
| Stage | MVP
| Not given
|
|---|
| Founded | 2026
| Not given
|
|---|
| Based in | GB
| Not given
|
|---|
| Pricing model | Contact for pricing
| Open source
|
|---|
| Pricing | Pricing tiers planned at launch: Basic, Pro, and Max for home users; Studio tier for developers; and per-seat Enterprise plans for organisations. Specific prices to be revealed at launch.
| Free and open source. No subscription. Pay only your own LLM provider costs, or nothing if you use a local model with Ollama.
|
|---|
| Key features | - Real-time voice recognition that knows who is talking
- Persistent memory across devices and conversations
- Contextual understanding that shapes tone and timing
- Desktop and browser control with task automation
- Permission-based actions with audit trail
- Overlord Mode for consulting multiple expert models
- Runs on Windows and macOS with mobile companion
| - 70+ LLM provider support including Ollama, OpenAI, Anthropic, and OpenAI-compatible endpoints
- 30+ communication channels including Telegram, Discord, WhatsApp, Slack, Signal, iMessage, email, and Matrix
- Supervised autonomy with approval-required and blocked action modes
- OS-level sandboxing with Landlock, Bubblewrap, Seatbelt, or Docker
- Cryptographic tool receipts with HMAC-SHA256 signing
- Single native Rust binary with no runtime dependencies
- Millisecond cold start in approximately 12 milliseconds on Apple Silicon
- Command allowlists and workspace scoping to control what runs
- Runs on ARM and x86 from Raspberry Pi to cloud VMs
- Memory footprint less than a browser tab
|
|---|
| What makes it different | - Purpose-built emotional computing platform, not a wrapper around someone else's model
- Deepens over time through continued interaction and learning
- Privacy-first design with data held in UK, never sold or used for profiling
- Built by solo founder with conviction-driven decisions rather than committee
- Presence that maintains continuity, not a stateless chatbot
| - Fully owned on your machine with your own keys or local models, no cloud seat or subscription
- Security by default with supervised autonomy, sandboxing, and cryptographic receipts rather than as an add-on
- Lightweight native binary architecture with millisecond startup and minimal memory overhead
- 70+ LLM providers and 30+ channels in a single deployable binary
- Open source under MIT or Apache-2.0 with active development and monthly releases
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | Not checked yet ● no change since last check | 57 ● no change since last check |
|---|
| Trust Flow (Majestic) | 0 | 10 |
|---|
| Citation Flow (Majestic) | 11 | 43 |
|---|
| Referring domains (Majestic) | 8 | 345 |
|---|
Morgan™
ZeroClaw