Jarvis
Authorisation required
Jarvis
Personal AI System — Initial Setup
Step 1 of 3
What should JARVIS call you?
Step 2 of 3
Choose your AI provider
JARVIS works with any of these. Your key is stored locally and sent only to the provider you choose.
OpenRouter
Access 100+ models including GPT-4, Claude, Gemini, Llama
Recommended
OpenAI
GPT-4o, GPT-4 Turbo, o1 models
Paid
Anthropic
Claude 3.5 Sonnet, Claude 3 Opus
Paid
Google Gemini
Gemini 2.0 Flash, Gemini 1.5 Pro
Free tier
Step 3 of 3
Paste your API key
Get your free key at openrouter.ai/keys — takes 30 seconds. Free tier includes access to powerful open-source models.
Your API key never leaves your device except to the AI provider you chose.
You can change it anytime in Settings. JARVIS does not store your key on any server.
Standby
--:--:--
JARVIS v3.2.1 AUDIO FIX
Audio Settings
Workshop
Themes
Goals & OKRs
Dev Queue
Knowledge
Social
Settings
Log Out
System
CPU
–
MEM
–
NET
–
Memory Vault
–Memories
–Links
–Clusters
–KB
Session Cost
–Tokens
–Est. Cost
–Turns
–Model
Evolution Feed
Live Intel BBC Tech
BTC––
ETH––
Weather–
Build Mode Active
System Status
CPU
5%
MEM
46%
NET
13%
STORAGE — 2.1 TB / 4 TB
HEALTH
All systems operational
Task Queue
Analyse competitor reportIn Progress
Prepare Q2 presentationPending
Review product requirementsPending
Client call — Acme CorpUpcoming
Browser
Search or enter URL
Search
Docs
Drive
Notion
GitHub
Reminders
Team stand-up
Daily 9:30 AM
Project review
May 23, 11 AM
Submit budget report
May 24, 5 PM
Automation
Data sync62%
Report generationCompleted
BackupCompleted
Live Tools
Code
Data
Search
Image
Files & Activity
Q2_Roadmap.pdf10:21
Market_Analysis.xlsxYesterday
Product_Requirements.docxMay 19
Design_System.figMay 18
Audio Settings
INPUT DEVICE (MICROPHONE)
Ready
OUTPUT DEVICE (SPEAKER)
System
Tasks
Browser
Reminders
Automation
Live Tools
Files
Audio
Standby
● Focus 25:00
AUDIO BOOT REPORT
Mic permission: checking…
Selected mic: checking…
Selected speaker: checking…
STT: checking…
TTS: checking…
setSinkId supported: checking…
Frontend version: v3.2.1
Conversation
Clear
MIC
WAKE
STT
LLM
TTS
SPK
PC
—
DEV
Task Timeline ✕
No active tasks
Developer Console ✕
boot Jarvis loaded
Tap orb to speak · Hold for memory map · Press ? for shortcuts
Interface Theme
Mode
Cyber (Default)
Crimson
Gold
Violet
Matrix
Dark (OLED)
Custom colour

⬡ Drop File

Release to send to JARVIS

☀ Morning briefing ready — tap to hear it
Keyboard Shortcuts
/ or SpaceFocus input
Ctrl+MToggle mic
Ctrl+BMemory map
Ctrl+WWorkshop
Ctrl+SSocial panel
Ctrl+KClear chat
Ctrl+IGenerate image
Ctrl+Shift+PShare screen
Ctrl+NQuick notes
Ctrl+GGoals & OKRs
Ctrl+QDev queue
Ctrl+UKnowledge wiki
EscClose panels
Press ? or Esc to close
Capabilities
Connect Social
Memory Vault
Share Screen
Quick Notes
Ctrl+N to open · saves to Obsidian vault
⬡ Analysing Video
Workshop
No active build detected
✕
Waiting for build context…
Neural Memory Map
Loading vault…
Vault Graph Knowledge Graph
–memories
–connections
–clusters
–kb stored
Goals & OKRs
Obsidian-backed objectives
Loading…
Dev Queue
Tasks JARVIS extracted from your conversations
Loading…
Knowledge Wiki
JARVIS's living profile of you — synced to Obsidian
Select a page above
Voice Output
Settings
Audio Input (Microphone)
Click to test
Audio Output (Speaker)
Choose which speaker Jarvis plays voice responses through.
Wake Word Status
Checking…
Engine: Browser SpeechRecognition (fallback). Not Deepgram.
Local Model (Ollama)
Scan to find Ollama models on your PC. Jarvis uses local first, falls back to cloud if it fails or is slow.
Voice Pipeline Tests
Run tests to verify pipeline
PC Bridge Status
Checking…
Audio Diagnostics
AI Provider
Leave blank to use JARVIS server defaults. Keys stored in localStorage only.
Your Profile
Reset