docs: document llmman as a local model server

docs/agents/models.md

@@ -14,10 +14,21 @@ The `model_args` parameter allows you to pass additional arguments to the model,
 
 The `api_base` parameter allows you to specify a custom API endpoint. This is useful when:
 
-- Using a local model server (e.g., Ollama, llama.cpp, llamafile)
+- Using a local model server (e.g., Ollama, llmman, llama.cpp, llamafile)
 - Routing through a proxy
 - Using a self-hosted model endpoint
 
+For example, [llmman](https://github.com/llmmanorg/llmman) serves the Ollama API on port `17434` instead of `11434`, so it can be used through the `ollama/` provider by pointing `api_base` at it:
+
+```python
+from any_agent import AgentConfig
+
+config = AgentConfig(
+    model_id="ollama/gemma4",
+    api_base="http://localhost:17434",
+)
+```
+
 The `api_key` parameter allows you to explicitly specify an API key for authentication. By default, `any-llm` will automatically search for common environment variables (like `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, etc.).
 
 See the [AnyLLM Provider Documentation](https://docs.mozilla.ai/any-llm/providers/) for the complete list of supported providers.

docs/cookbook/agent-with-local-llm.md

@@ -37,6 +37,9 @@ we will use [Ollama](https://ollama.com/) to run our LLM locally!
 
 First, install Ollama by following their instructions: https://ollama.com/download
 
+>**Note**: [llmman](https://github.com/llmmanorg/llmman) serves the same Ollama API on port `17434`.
+> To use it instead, keep the `ollama/` prefix in `model_id` and add `api_base="http://localhost:17434"` to your `AgentConfig`.
+
 ### Picking an LLM
 
 Pick a model that you can run locally based on your hardware and download it from your terminal. For example:

docs/cookbook/agent_with_local_llm.ipynb

@@ -64,6 +64,9 @@
     "\n",
     "First, install Ollama by following their instructions: https://ollama.com/download\n",
     "\n",
+    ">**Note**: [llmman](https://github.com/llmmanorg/llmman) serves the same Ollama API on port `17434`.\n",
+    "> To use it instead, keep the `ollama/` prefix in `model_id` and add `api_base=\"http://localhost:17434\"` to your `AgentConfig`.\n",
+    "\n",
     "### Picking an LLM\n",
     "\n",
     "Pick a model that you can run locally based on your hardware and download it from your terminal. For example:\n",

src/any_agent/config.py

@@ -86,8 +86,8 @@ class AgentConfig(BaseModel):
     api_base: str | None = None
     """Custom API endpoint URL for the model provider.
 
-    Use this to specify custom endpoints for local models (Ollama, llama.cpp, etc.) or proxy services.
-    For example: `http://localhost:11434/v1` for Ollama.
+    Use this to specify custom endpoints for local models (Ollama, llmman, llama.cpp, etc.) or proxy services.
+    For example: `http://localhost:11434/v1` for Ollama, or `http://localhost:17434` for llmman.
     """
 
     api_key: str | None = None