docs: Add ollama configuration example

[johnsaigle]

Tested this locally and got Ollama to work via opencode run "test prompt".

The "Requirements" section describe the current behaviour of the program, but these might in fact be bugs, (see #227). I'm happy to update the PR if these instructions aren't correct.

[semidark]

If Opencode uses ai-sdk, as stated here in the documentation, shouldn’t we be able to use the Ollama API directly, without the need for the OpenAI API layer?

See: https://ai-sdk.dev/providers/community-providers/ollama

I tried it with the following example (see ollama part of the config), but it is not yet working. Only the ollama-v1 part of the config is somehow working.

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "ollama-v1": {
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "http://ollama.lan:11434/v1"
      },
      "models": {
        "qwen3-32b:MAX": {
	  "name"      : "Qwen3 32b - Max",
	  "tool_call" : true,
	  "reasoning" : true,
	  "limit": {
	    "context": 22000,
	    "output": 8192
	  } 
	},
        "qwen3-14b:MAX": {
	  "name"      : "Qwen3 14b - Max",
	  "tool_call" : true,
	  "reasoning" : true,
	  "limit": {
	    "context": 40960,
	    "output": 8192
	  } 
	}
      }
    },
    "ollama": {
      "npm": "@ai-sdk/ollama-ai-provider",
      "options": {
        "baseURL": "http://ollama.lan:11434"
      },
      "models": {
        "qwen3-32b:MAX": {
          "name": "Qwen3 32b - Max"
	}
      }
    }
  }
}

[johnsaigle]

Closing this because it's probably very stale