[FEATURE]:Interactive configuration for local/self-hosted LLMs #49319

Feature hasn't been suggested before.

• I have verified this feature I'm about to request hasn't been suggested before.

Describe the enhancement you want to request

Problem

Configuring a local or self-hosted LLM in OpenCode currently requires manually editing the OpenCode configuration file.

For users who want to run models locally, this creates unnecessary setup friction. Users have to understand the configuration structure and manually enter the provider, endpoint, model name, and other options before they can use the model.

Proposed solution

Add an interactive setup flow in the TUI/CLI for local and self-hosted models.

For example:

Select provider type:

Local / Self-hosted LLM
Cloud Provider
...

Then:

Local provider:

Endpoint:

http://localhost:11434/v1

Model:

Enter model name

API key:

Optional

Protocol:

OpenAI-compatible

OpenCode could optionally query the endpoint for available models and allow the user to select one instead of manually entering the model name.

For example:

Available models:

qwen2.5-coder:3b
qwen2.5-coder:1.5b
llama-3.1-8b

After confirmation, OpenCode could automatically generate/update the corresponding configuration.

Why this would be useful

This would make local LLM setup much easier for users who:

Run models locally
Use self-hosted inference servers
Frequently switch between local models
Use OpenAI-compatible local APIs
Are not familiar with OpenCode's configuration format

The feature should not be limited to Ollama. It should work with any supported local/self-hosted endpoint.

Suggested UX

A command such as:

/config

or:

/model add

could provide an interactive configuration wizard:

Provider type
↓
Local / Self-hosted
↓
Endpoint
↓
Detect available models
↓
Select model
↓
Configure optional settings
↓
Save

The existing configuration-file functionality could remain available for advanced users.

Expected result

Users should be able to add a local model without manually opening or editing opencode.json.

Advanced users could continue using the configuration file, while new users would have a simpler interactive setup path.

Scope

This request is specifically about improving the configuration experience for local/self-hosted LLMs in general, rather than adding support for a particular provider.