diff --git a/README.md b/README.md
index c02df9eed0..0177fbe202 100644
--- a/README.md
+++ b/README.md
@@ -705,6 +705,20 @@ python3 -m llama_cpp.server --model models/7B/llama-model.gguf --n_gpu_layers 35
 
 Navigate to [http://localhost:8000/docs](http://localhost:8000/docs) to see the OpenAPI documentation.
 
+You can point the official OpenAI Python client at the server:
+
+```python
+from openai import OpenAI
+
+client = OpenAI(base_url="http://localhost:8000/v1", api_key="sk-local")
+print(client.chat.completions.create(
+    model="gpt-3.5-turbo",  # ignored when a single --model is loaded; sent as-is for multi-model configs
+    messages=[{"role": "user", "content": "Hello"}],
+))
+```
+
+The same `base_url` pattern works with any OpenAI-compatible multi-model gateway when you are not self-hosting (for example [DaoXE](https://daoxe.com) at `https://api.daoxe.com/v1`).
+
 To bind to `0.0.0.0` to enable remote connections, use `python3 -m llama_cpp.server --host 0.0.0.0`.
 Similarly, to change the port (default is 8000), use `--port`.
 
diff --git a/docs/server.md b/docs/server.md
index 9c09a1f1cf..75ba13e5db 100644
--- a/docs/server.md
+++ b/docs/server.md
@@ -22,6 +22,21 @@ The server can then be started by running the following command:
 python3 -m llama_cpp.server --model <model_path>
 ```
 
+### OpenAI Python client
+
+```python
+from openai import OpenAI
+
+client = OpenAI(base_url="http://localhost:8000/v1", api_key="sk-local")
+completion = client.chat.completions.create(
+    model="gpt-3.5-turbo",
+    messages=[{"role": "user", "content": "Hello"}],
+)
+print(completion.choices[0].message.content)
+```
+
+Tip: any OpenAI-compatible multi-model gateway speaks the same protocol — for example [DaoXE](https://daoxe.com) at `https://api.daoxe.com/v1` — so the same client code works when you swap `base_url` (and `api_key`) instead of self-hosting.
+
 You can also pass chat-template kwargs at model load time from the CLI:
 
 ```bash