Your own Nexus and AI coding tools

Available on Standard and Enterprise.

On Pay as you Go, coding and generation use hub defaults. On Standard and Enterprise, you configure Workspace Settings → Nexus so AI coding follows your stack:

  • Point at your organization’s Nexus endpoint (Protected AI on Diode)
  • Choose default provider / model (including catalogs such as OpenRouter when configured)
  • Extend Nexus with access to organization data and tools so sessions are organization-aware — policy-driven generation grounded in your systems

That is the difference between “an LLM wrote some code” and “our approved coding infrastructure produced a change we can defend.” On the Nexus side, that stack is built around Private AI models — self-hosted or partner-hosted inference plus org tools.

In the product

Operators use Nexus from project Coding chat and (when enabled) Create-project auto-generate. Agents can also start generation with project/generate when the workspace API key allows it.

Enterprise sharing

Enterprise workspaces can share Nexus / AI coding configurations with connected workspaces and divisions — so a central platform team sets the approved stack once. See Share infrastructure across workspaces.