Loading...
How can I help you today?
Loading neural network...
Loading neural stats...
Loading models...
Checking security...
Multi-user AI assistant
Initial admin password is generated on first run and stored in ~/.ai_users/initial_admin_credentials.json.
This account was created with an auto-generated password. Please set your own password before continuing.
Upload a document for AI analysis
Open the Siren AI Android app and tap Scan QR Code to auto-configure your device for the cluster.
Install Siren AI on other devices or platforms. All installers connect back to this server.
Run this in PowerShell on your Windows machine:
irm /download/windows_ps1 | iex
Click to copy
⚠️ You're viewing this over HTTPS with Siren's self-signed certificate -- irm will refuse it by default with an SSL/TLS trust error. Easiest fix: open this page via the plain
http:// address on port 5050 instead (regenerates the command above with no cert issue). Otherwise, run one of these first in the same PowerShell window, depending on your version ($PSVersionTable.PSVersion to check):
• Windows PowerShell 5.1 (the Windows default):
[System.Net.ServicePointManager]::ServerCertificateValidationCallback = {$true}
• PowerShell 7+ (add the flag directly to the install command instead):
irm -SkipCertificateCheck /download/windows_ps1 | iex
Generate a one-time code, then paste it into the agent when it asks after install/launch.
Enable Unknown Sources in Android Settings → Security, then tap the APK link above. Or scan the QR code from the Pair Device panel.
After uploading a new app-debug.apk to the downloads folder, publish its version here so paired phones get an automatic in-app update prompt (they poll periodically — no need to reconnect).
Currently published: —
| "Open Firefox" | Launch applications |
| "Search for [topic]" | Web search with live results |
| "Volume up/down" | System volume control |
| "What time is it?" | Time/date queries |
| "Take a screenshot" | Capture screen |
| "Lock screen" | System lock |
| "Shutdown / Restart" | Power control |
| "Send email to [name]" | Email compose via voice |
| "Read my emails" | Check & read inbox aloud |
| "Generate image of..." | AI image generation |
| "Create a document" | Start document creation |
| "Run command [cmd]" | Execute system commands |
| "Research [topic]" | Open Research Workspace |
| "Diagnose the system" | Run system health check |
| "Set a reminder..." | Create reminder via Personal Assistant |
| "Click at 500, 300" | Click at exact screen coordinates |
| "Click the save button" | Vision-based — locates the element on screen (best-effort estimate, not pixel-perfect) |
| "Double-click at 100, 200" | Double-click at coordinates |
| "Move the mouse to 400, 400" | Move cursor without clicking |
| "Scroll down" / "Scroll up 5" | Scroll the active window |
| "Type hello world" | Type text into the focused window |
| "Press Enter" / "Press ctrl+a" | Send a key or key combination |
| "Copy [text] to the clipboard" | Set the system clipboard |
| "What's in my clipboard?" | Read the system clipboard |
| "Open the URL example.com" | Open a link in your default browser |
| "Generate an image of..." | Creates an image from text description |
| "Create a video of..." | Video generation (GPU required) |
| "Analyze this image" | Upload an image for AI vision analysis (uses llava model) |
| Upload button (📄) | Attach files or images directly in chat |
| Settings → Image & Video | Configure resolution, steps, models |
| Code & debugging | → qwen2.5-coder (optimized for code) |
| Math & logic | → qwen2.5 (strong reasoning) |
| Image analysis | → llava (vision model) |
| General chat | → llama3.2 or selected model |
/api/system/updates for package data, /api/system/optimizations for hardware suggestions| General | AI name, wake word, theme (dark/light toggle in header), language |
| Voice & Audio | TTS engine, speed, pitch, voice browser, noise profile, voice training |
| LLM Models | Model selection, download, delete, auto-switch config |
| Image & Video | Generation model settings, resolution, steps |
| Neural & Learning | Training controls, dream cycles, brain dashboard |
| Cluster | Node management, benchmarks, resource monitoring |
| Security & Firewall | Scan, audit, CVE lookup, active defense, sudo approvals |
| Integrations | Email, calendar, *arr, Home Assistant, n8n, NZBGet |
| Avatar | AI persona, appearance, 3D avatar creator |
| System Health | Full system status report |
| Self-Repair | Auto-diagnose and fix issues |
| Neural Map | 3D neural network visualization |
| Image Gen | Generate image from text |
| Dream Cycle | Neural optimization (idle time) |
| Research Topic | Research Workspace — notebooks, sources, Q&A |
| Learn Topic | Deep research and store knowledge |
| Video Gen | Generate video (GPU required) |
| Neural Dashboard | Training stats and metrics |
| Model Manager | Download, switch, delete models |
| Security | Firewall scan, audit, CVE lookup, active defense |
| Cluster | Cluster health dashboard |
| 3D Avatar | 3D avatar creator page |
| Documents | Document manager — create, edit, export |
| Command DB | Browse learned command database |
| Pair Device | Pair Android companion |
| Jupyter | Open JupyterLab IDE |
| Inbox viewer with folders | |
| Media Library | Plex library scanner & cleaner |
| Media Coord | Radarr / Sonarr / Lidarr manager |
| Storage | Drive status, NZBGet, media services |
| Integrations | n8n, Home Assistant, calendar, email |
| Assistant | Reminders, tasks, calendar events |
| Voice Training | Speaker recognition & calibration |
| Diagnostics | System diagnostics & log viewer |
| Enter | Send message |
| Shift + Enter | New line in message |
| Ctrl + / | Toggle settings panel |
| Escape | Close panels/modals |
| http://host:5050 | Direct HTTP (local network) |
| https://host:5443 | HTTPS with self-signed certificate |
| http://host:5080 | Auto-redirects to HTTPS |