AI Model System
Nexus is a sovereign operating partner powered by your preferred language models. Bring your own keys (BYOK), select optimized AI Profiles, and control data residency entirely within your sandboxed environment.
Audience: Subscribers, founders, and administrators managing model connectivity and data residency.
Current Version
1.0
Documentation aligned to the current operational build of Chat, Automations, Contacts, and workflow runtime.
Scope
1
A standalone documentation page with a stable URL you can share directly.
Quick Start
How the system is meant to be used
Step 1
Read the platform overview first so you understand how Chat, apps, and runtime work together.
Step 2
Start your first conversation in Chat with a real business outcome, not just a general prompt.
Step 3
Let Nexus turn that intent into a durable record, workflow, or next-step artifact.
Step 4
Connect the systems required for context, communication, and execution as setup requires.
Step 5
Move into the relevant app surface to review, manage, approve, and operate the resulting work.
Operating Principle
Chat compiles. Apps operate. Runtime executes.
Nexus is not a chatbot that happens to call tools. It is a structured operating system where conversational intent is translated into durable records, validated state, and auditable actions.
This docs site is the product-level reference for how that model works today and how the current 1.0 surfaces fit together.
Audience
Subscribers, founders, and administrators managing model connectivity and data residency.
AI Model System
Nexus is a sovereign operating partner powered by your preferred language models. Bring your own keys (BYOK), select optimized AI Profiles, and control data residency entirely within your sandboxed environment.
Philosophy
AI capability should be a secure, private utility that you control. You shouldn't have to trade context or data privacy for intelligence; your keys, your profiles, and your models stay strictly inside your instance.
Technically Managed By
- Bring Your Own Key (BYOK): Secure connection of Google, OpenAI, Anthropic, or OpenRouter.
- Sovereign Sandboxing: Dynamic credential injection ensures keys never leave your secure host.
- AI Profiles: Purpose-built operating profiles mapping model selection to active goals.
- Auto-Refreshed OAuth: Real-time background token sync prevents session expiration.
How Users Should Use and Manage It
- 1Navigate to Settings > AI Configurations to connect your preferred provider keys.
- 2Select a designated AI Profile (Quality, Speed, Coding) to tune the system's defaults.
- 3Enable Advanced Override to select specific individual models for advanced streams.
How It Fits in Nexus
- The AI Model System is the cognitive engine of your Nexus instance.
- It coordinates with OpenClaw to safely run reasoning chains, tools, and actions.
- It respects your billing tier and feature gates to keep operations governed and predictable.
What Matters
- Dynamic key injection: Secrets are sandboxed securely on your private server.
- Intelligent AI Profiles: Fast, Balanced, or Quality optimized presets.
- Revocation & Safety: Total control to connect, rotate, or completely wipe credentials instantly.
Reference Notes
Next Step
Use the docs as the product contract.
As new apps and runtime behaviors ship, extend this documentation hub first so users and internal teams have one clear reference for how Nexus is supposed to work.