Add AnyLLMClient as alternative to LiteLLMClient

README.md

@@ -47,6 +47,7 @@ pip install 'stirrup[all]'  # or: uv add 'stirrup[all]'
 
 # Individual extras
 pip install 'stirrup[litellm]'  # or: uv add 'stirrup[litellm]'
+pip install 'stirrup[anyllm]'   # or: uv add 'stirrup[anyllm]'
 pip install 'stirrup[docker]'   # or: uv add 'stirrup[docker]'
 pip install 'stirrup[e2b]'      # or: uv add 'stirrup[e2b]'
 pip install 'stirrup[mcp]'      # or: uv add 'stirrup[mcp]'
@@ -164,6 +165,27 @@ agent = Agent(
 
 See [LiteLLM Example](https://stirrup.artificialanalysis.ai/examples/#litellm-multi-provider-support) or [Deepseek Example](https://stirrup.artificialanalysis.ai/examples/#openai-compatible-apis-deepseek-vllm-ollama) for complete examples.
 
+### any-llm (Anthropic, Google, etc.)
+
+```python
+# Ensure any-llm is added with: pip install 'stirrup[anyllm]'  # or: uv add 'stirrup[anyllm]'
+# Create any-llm client for Anthropic Claude
+# See https://mozilla-ai.github.io/any-llm/providers/ for all supported providers
+client = AnyLLMClient(
+    model_slug="anthropic/claude-opus-5",
+    max_tokens=8_192,
+    context_window_tokens=1_000_000,
+)
+
+# Pass client to Agent - model info comes from client.model_slug
+agent = Agent(
+    client=client,
+    name="claude_agent",
+)
+```
+
+See [any-llm Example](https://stirrup.artificialanalysis.ai/examples/#any-llm-multi-provider-support) or [Deepseek Example](https://stirrup.artificialanalysis.ai/examples/#openai-compatible-apis-deepseek-vllm-ollama) for complete examples.
+
 ## Default Tools
 
 When you create an `Agent` without specifying tools, it uses `default_tools()`:

docs/api/clients/anyllm.md

@@ -0,0 +1,12 @@
+# any-llm Client
+
+!!! note "Optional Dependency"
+    any-llm is an optional dependency. Install with:
+    ```bash
+    pip install stirrup[anyllm]  # or: uv add stirrup[anyllm]
+    ```
+
+`max_tokens` limits provider output. `context_window_tokens` (required) separately
+tells the agent when conversation history should be summarized.
+
+::: stirrup.clients.anyllm_client

docs/concepts.md

@@ -268,6 +268,25 @@ Use `LiteLLMClient` for Anthropic, Google, and other providers via [LiteLLM](htt
 !!! note "LiteLLM Installation"
     Requires `pip install stirrup[litellm]` (or: `uv add stirrup[litellm]`)
 
+### AnyLLMClient
+
+Use `AnyLLMClient` for Anthropic, Google, and other providers via [any-llm](https://mozilla-ai.github.io/any-llm/providers/):
+
+```python
+--8<-- "examples/anyllm_example.py:client"
+```
+
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| `model_slug` | `str` | required | Provider/model string (e.g., `"anthropic/claude-opus-5"`) |
+| `max_tokens` | `int` | `64_000` | Maximum provider response/output tokens |
+| `context_window_tokens` | `int` | required | Context capacity for summarization |
+| `reasoning_effort` | `str \| None` | `None` | For reasoning models (e.g. GPT-5.6) |
+| `kwargs` | `dict \| None` | `None` | Additional provider-specific arguments |
+
+!!! note "any-llm Installation"
+    Requires `pip install stirrup[anyllm]` (or: `uv add stirrup[anyllm]`)
+
 ### Creating Your Own Client
 
 Implement the `LLMClient` protocol to create a custom client:

docs/examples.md

@@ -98,6 +98,17 @@ Use LiteLLM to connect to non-OpenAI providers like Anthropic Claude, Google Gem
 !!! note
     Requires `pip install stirrup[litellm]` (or: `uv add stirrup[litellm]`) and the appropriate API key for your chosen provider (e.g., `ANTHROPIC_API_KEY` for Claude).
 
+## any-llm Multi-Provider Support
+
+Use any-llm to connect to non-OpenAI providers like Anthropic Claude, Google Gemini, and many others.
+
+```python
+--8<-- "examples/anyllm_example.py"
+```
+
+!!! note
+    Requires `pip install stirrup[anyllm]` (or: `uv add stirrup[anyllm]`) and the appropriate API key for your chosen provider (e.g., `ANTHROPIC_API_KEY` for Claude).
+
 ## Custom Finish Tool
 
 Define structured output with a custom finish tool:

docs/extending/clients.md

@@ -4,13 +4,14 @@ This guide covers implementing custom LLM clients for Stirrup.
 
 ## Built-in Clients
 
-Stirrup provides three built-in clients:
+Stirrup provides four built-in clients:
 
 | Client | API | Best For |
 |--------|-----|----------|
 | [`ChatCompletionsClient`](../api/clients/chat_completions.md) | OpenAI Chat Completions | OpenAI, OpenRouter, vLLM, Ollama, and other OpenAI-compatible APIs |
 | [`OpenResponsesClient`](../api/clients/open_responses.md) | OpenAI Responses API | Providers implementing the newer Responses API format |
 | [`LiteLLMClient`](../api/clients/litellm.md) | LiteLLM | Multi-provider support (Anthropic, Google, Azure, etc.) |
+| [`AnyLLMClient`](../api/clients/anyllm.md) | any-llm | Multi-provider support (Anthropic, Google, etc.) |
 
 ## LLMClient Protocol
 

docs/getting-started.md

@@ -26,6 +26,7 @@ pip install 'stirrup[all]'  # or: uv add 'stirrup[all]'
 | Extra | Description |
 |-------|-------------|
 | `litellm` | Multi-provider support via LiteLLM (Anthropic, Google, etc.) |
+| `anyllm` | Multi-provider support via any-llm (Anthropic, Google, etc.) |
 | `docker` | Docker-based code execution |
 | `e2b` | E2B cloud sandboxes |
 | `mcp` | MCP server support |

docs/index.md

@@ -30,6 +30,7 @@ pip install 'stirrup[all]'  # or: uv add 'stirrup[all]'
 
 # Individual extras
 pip install 'stirrup[litellm]'  # or: uv add 'stirrup[litellm]'
+pip install 'stirrup[anyllm]'   # or: uv add 'stirrup[anyllm]'
 pip install 'stirrup[docker]'   # or: uv add 'stirrup[docker]'
 pip install 'stirrup[e2b]'      # or: uv add 'stirrup[e2b]'
 pip install 'stirrup[mcp]'      # or: uv add 'stirrup[mcp]'
@@ -73,6 +74,14 @@ For non-OpenAI providers, change the base URL of the `ChatCompletionsClient`, us
 
 See [LiteLLM Example](examples.md#litellm-multi-provider-support) or [Deepseek Example](examples.md#openai-compatible-apis) for complete examples.
 
+### any-llm (Anthropic, Google, etc.)
+
+```python
+--8<-- "examples/anyllm_example.py:client"
+```
+
+See [any-llm Example](examples.md#any-llm-multi-provider-support) or [Deepseek Example](examples.md#openai-compatible-apis) for complete examples.
+
 ## Default Tools
 
 When you create an `Agent` without specifying tools, it uses `default_tools()`:

examples/anyllm_example.py

@@ -0,0 +1,43 @@
+"""Example: Using Agent with any-llm for multi-provider support.
+
+Demonstrates using any-llm to connect to non-OpenAI providers like Anthropic Claude,
+Google Gemini, etc. For any-llm, create the client directly and pass it to the Agent.
+
+Requires:
+- uv pip install stirrup[anyllm]
+- ANTHROPIC_API_KEY environment variable (for this example)
+"""
+
+import asyncio
+
+from stirrup import Agent
+from stirrup.clients.anyllm_client import AnyLLMClient
+
+
+async def main() -> None:
+    """Run an agent using Anthropic Claude via any-llm."""
+    # --8<-- [start:client]
+    # Create any-llm client for Anthropic Claude
+    # See https://mozilla-ai.github.io/any-llm/providers/ for all supported providers
+    client = AnyLLMClient(
+        model_slug="anthropic/claude-opus-5",
+        max_tokens=8_192,
+        context_window_tokens=1_000_000,
+    )
+
+    # Pass client to Agent - model info comes from client.model_slug
+    agent = Agent(
+        client=client,
+        name="claude_agent",
+    )
+    # --8<-- [end:client]
+
+    async with agent.session(output_dir="./output/anyllm_example") as session:
+        await session.run(
+            "What has the temperature been in the last 3 days in San Francisco? "
+            "Provide a brief summary and output a pdf file with the summary and a graph."
+        )
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

mkdocs.yml

@@ -103,6 +103,7 @@ nav:
           - ChatCompletions: api/clients/chat_completions.md
           - OpenResponses: api/clients/open_responses.md
           - LiteLLM: api/clients/litellm.md
+          - any-llm: api/clients/anyllm.md
       - Tools:
           - Overview: api/tools/index.md
           - Web Tools: api/tools/web.md

pyproject.toml

@@ -39,13 +39,14 @@ dependencies = [
 
 [project.optional-dependencies]
 litellm = ["litellm>=1.79.3"]
+anyllm = ["any-llm-sdk[gemini]>=1.27.0,<1.29.0"]
 e2b = ["e2b-code-interpreter>=2.3.0"]
 e2b-throttle = ["stirrup[e2b]", "aiolimiter>=1.1.0"]
 docker = ["docker>=7.0.0", "python-dotenv>=1.0.0"]
 mcp = ["mcp>=1.9.0"]
 browser = ["browser-use>=0.11.3"]
 slack = ["slack-bolt>=1.20.0"]
-all = ["stirrup[litellm,e2b,e2b-throttle,docker,mcp,browser,slack]"]
+all = ["stirrup[litellm,anyllm,e2b,e2b-throttle,docker,mcp,browser,slack]"]
 
 [project.urls]
 Homepage = "https://github.com/ArtificialAnalysis/Stirrup"

src/stirrup/clients/__init__.py

@@ -8,6 +8,9 @@
 
 For multi-provider support via LiteLLM, install the litellm extra:
     pip install stirrup[litellm]
+
+For multi-provider support via any-llm, install the anyllm extra:
+    pip install stirrup[anyllm]
 """
 
 from stirrup.clients.chat_completions_client import ChatCompletionsClient