Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.stagewise.io/llms.txt

Use this file to discover all available pages before exploring further.

stagewise ships with a curated list of built-in models. Add custom models for anything not in the default list — fine-tuned variants, self-hosted models, or models from alternative providers.

​
Adding a custom model

1

Open Models & Providers

Go to Settings → Agent → Models & Providers and scroll to the Custom Models section.
2

Click Add model

The model editor opens.
3

Configure the model

Provide:
  • Model ID — The identifier your endpoint expects
  • Display name — How the model appears in the selector
  • Provider — Which built-in provider this model routes through
The endpoint mode configured for that provider determines where requests go.

​
Required capabilities

For a model to work in stagewise, it must support:
  • Text generation — Standard text output
  • Tool calling — Structured tool use via the API
Models lacking either capability fail at runtime.

​
Optional capabilities

Not required but enhance the agent:
  • Image vision — Understand image inputs
  • Video vision — Understand video inputs
  • Audio — Process audio input
stagewise does not require native file input support. Files are read through a custom pipeline independent of the model’s file handling.

​
Custom model with a local endpoint

A typical workflow for self-hosted models:
  1. Create a custom endpoint pointing to your server (e.g., http://localhost:11434/v1 for Ollama)
  2. Create a custom model with the model ID your server expects
  3. Select the custom model from the model picker in the chat sidebar

​
Provider options and headers

For advanced configurations:
  • Provider options — Raw JSON passed to the AI SDK for provider-specific behavior (reasoning effort, thinking budgets, streaming options)
  • Custom headers — Additional HTTP headers sent with every request to this model