diff --git a/QUICKSTART.md b/QUICKSTART.md
index 52e5216e..2d35899e 100644
--- a/QUICKSTART.md
+++ b/QUICKSTART.md
@@ -257,6 +257,28 @@ degrading to a different search method.
 You can replace OpenRouter with another OpenAI-compatible LLM endpoint by
 changing the `[llm]` model, base URL, and key.
 
+## Use another OpenAI-compatible endpoint
+
+The walkthrough above uses OpenRouter, but the same `[llm]` settings work
+with any OpenAI-compatible endpoint. For example, to use the official
+DeepSeek API, set these three values in `~/.everos/everos.toml`:
+
+```toml
+[llm]
+model = "deepseek-flash"
+api_key = "<DEEPSEEK_API_KEY>"
+base_url = "https://api.deepseek.com"
+```
+
+Then start the server with `everos server start`. This is a config-only
+change: no EverOS code changes are required. To switch models, update
+`model` to another identifier offered by your endpoint and restart the
+server.
+
+The same pattern applies to any OpenAI-compatible provider: point
+`base_url` at the provider's endpoint, set `model` to the provider's model
+identifier, and put the provider's key in `api_key`.
+
 ## Stop the server
 
 Press `Ctrl+C` in the server terminal.
diff --git a/docs/configuration.md b/docs/configuration.md
index 5cf85a19..485b37fe 100644
--- a/docs/configuration.md
+++ b/docs/configuration.md
@@ -124,6 +124,10 @@ Zilliz Cloud endpoint; a Milvus Lite filesystem path is rejected.
 
 ### `[llm]`
 
+Any OpenAI-compatible endpoint is supported: set `base_url` to the
+provider's endpoint (for example `https://api.deepseek.com` with
+`model = "deepseek-flash"`) and `api_key` to that provider's key.
+
 | Field | Type | Default | Required | Description |
 |---|---|---|---|---|
 | `model` | string | `"gpt-4.1-mini"` | No | LLM model identifier. |