| What it does | Frontier Red Team Lab for LLM Safety
| AI assistant for Mac and Windows that lives in your menu bar or system tray, sees your screen, hears your voice, and helps get work done
|
|---|
| Stage | Startup
| Not given
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | Not given
| SE
|
|---|
| Pricing model | Freemium
| Freemium
|
|---|
| Pricing | Free: $0 with 30,000 daily tokens; Plus: $5.99/month with 80,000 daily tokens; Standard: $19.99/month with 265,000 daily tokens; Pro: $49.99/month with 665,000 daily tokens. Cryptocurrency payment offers 2 months access for 1 month payment.
| Free: 30,300 tokens every month. Top-up tokens: $5 per million tokens. Individual bundles: $2 (for 1M tokens), $5, $10, or $25. Buy once or set up automatic monthly top-ups. Tokens never expire. Referral discount: 5% off for referred friends and 5% off for you per friend who buys, up to 25% total.
|
|---|
| Key features | - DarkEval Benchmark - proprietary adversarial evaluation system with 96% exploit coverage
- Memory - context-aware conversations saving approximately 40% on token usage
- Web search capability across platform
- Transparent reasoning and chain-of-thought visibility
- Compliance mapping to NIST AI RMF, ISO/IEC 23894, and SOC 2
- Tool calling and JSON output support
- Multiple open-weight frontier models available
- Enterprise-grade reporting and team analytics
| - Sees your screen and answers questions about what's displayed
- Talks you through tasks out loud and points at where to click
- Runs applications, plays music, sorts files, manages Calendar events and GitHub issues
- Dictates text into any field and rewrites selected text
- Saves multi-step tasks as agents for reuse
- Remembers information between conversations on your device
- Focus on circled parts of the screen
- Reads and answers questions about PDFs, images, documents and code files
- On-device meeting transcription with summary on Mac
- Transfers files securely between your Mac and Windows PC over your network
|
|---|
| What makes it different | - Constraints-relaxed models uniquely capable of detecting vulnerabilities other filtered AI systems cannot
- Sandboxed environment ensuring all testing is isolated, logged, and auditable
- Purpose-built for professional red-teaming with enterprise compliance capabilities
- Free tier available with no credit card required for getting started
| - Voice-first interface: hold a key and speak naturally, no commands to learn
- Sees your screen so you can ask about anything displayed without describing it
- Points at actions instead of explaining them, keeping you in control
- Stores memory and files on your device, not the cloud
- Teaches you how to do things rather than replacing your actions
|
|---|
| Free plan | Not given
| Not given
|
|---|
| Open source | Not given
| Not given
|
|---|
| Platforms | Not given
| Not given
|
|---|
| Public API | Not given
| Not given
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | 42 ▼ −1 (down) since last check | 1 ● no change since last check |
|---|
| Trust Flow (Majestic) | 19 | 0 |
|---|
| Citation Flow (Majestic) | 33 | 17 |
|---|
| Referring domains (Majestic) | 382 | 4 |
|---|
Shannon AI