Skip to content
Devora

Features · Model catalogue

The right model for the moment.

Your team connects OpenAI-compatible providers in the owner console. The app lists exactly the models enabled for you, with a picker per conversation and named model roles for different jobs.

modelsenabled for you
ModelProviderAccess
atlas-largeyour endpointEnabled
atlas-miniyour endpointEnabled
legacy-chat—Off
The picker only lists models your account can use

Interface illustration — sample catalogue.

What it does

01

Your providers

Any OpenAI-compatible endpoint is configured once by your team — base URL, key and retries.

02

A clear catalogue

Models are discovered and enabled server-side; the app shows only what you can actually use.

03

Per-conversation choice

Switch models mid-work, with named model roles you can assign.

Details

What you can rely on.

The small print, written plainly — including the parts that depend on your account.

  • Model capabilities are set per model in the owner console.
  • The picker reflects your permissions; disabled models never appear.
  • Default and fallback models are configured by your team, not hardcoded in the app.

Availability

Model access depends on your plan and the models your team enables in the owner console.

Keep exploring.

Try Devora.

Install the desktop app, open your project, and start with a question about your own codebase.

Windows 10 & 11 · Local workspace · Reviewable edits