diff --git a/README.md b/README.md
index 51574b4..2153e30 100644
--- a/README.md
+++ b/README.md
@@ -170,6 +170,8 @@ Free-Way exposes **both** OpenAI and Anthropic compatible endpoints, so most cod
 - [OpenRouter Provider Setup](./docs/providers/openrouter.md)
 - [OpenCode setup guide](./docs/agents/opencode.md)
 - [Aider Integration Guide](./docs/agents/aider.md)
+- [Continue.dev setup guide](./docs/agents/continue.md)
+- [Cline / Roo Code setup guide](./docs/agents/cline-roo-code.md)
 
 ### Claude Code
 
diff --git a/docs/agents/cline-roo-code.md b/docs/agents/cline-roo-code.md
new file mode 100644
index 0000000..2defb82
--- /dev/null
+++ b/docs/agents/cline-roo-code.md
@@ -0,0 +1,34 @@
+# Cline / Roo Code Setup Guide
+
+Follow these steps to connect Cline or Roo Code to Free-Way through the OpenAI-compatible endpoint.
+
+## 1. Choose an OpenAI-compatible provider mode
+
+In Cline or Roo Code, choose the OpenAI-compatible provider option and set:
+
+```text
+Base URL: http://localhost:8787/v1
+API key: your FREEWAY_API_KEY
+Model: llama-3.3-70b
+```
+
+Use a model ID that appears in your local Free-Way model catalog. If Free-Way gateway auth is disabled, the API key can be any non-empty placeholder value.
+
+## 2. Keep the path shape exact
+
+Use `http://localhost:8787/v1` for OpenAI-compatible clients.
+
+Do not use `http://localhost:8787` in OpenAI-compatible provider mode, because the client will usually append `/chat/completions` or `/models` under `/v1`.
+
+## 3. Verify the connection
+
+Send a small prompt from Cline or Roo Code, then check Free-Way's **Usage** tab to confirm the request reached your local gateway.
+
+## Troubleshooting
+
+| Problem | Check |
+| --- | --- |
+| 404 or route not found | The base URL includes `/v1` |
+| Model not found | The model ID exists in Free-Way's model catalog |
+| Unauthorized | `FREEWAY_API_KEY` matches the gateway key configured in Free-Way |
+| No request in Usage | The selected provider mode is OpenAI-compatible and points to localhost |
diff --git a/docs/agents/continue.md b/docs/agents/continue.md
new file mode 100644
index 0000000..d2dadb3
--- /dev/null
+++ b/docs/agents/continue.md
@@ -0,0 +1,46 @@
+# Continue.dev Setup Guide
+
+Follow these steps to connect Continue.dev to Free-Way through the OpenAI-compatible endpoint.
+
+## 1. Confirm the Free-Way endpoint
+
+OpenAI-compatible clients should use:
+
+```text
+http://localhost:8787/v1
+```
+
+Anthropic-compatible clients should use `http://localhost:8787` without `/v1`.
+
+## 2. Add a model to Continue
+
+Add this entry to your Continue `config.json`:
+
+```json
+{
+  "models": [
+    {
+      "title": "Free-Way",
+      "provider": "openai",
+      "model": "llama-3.3-70b",
+      "apiBase": "http://localhost:8787/v1",
+      "apiKey": "your FREEWAY_API_KEY"
+    }
+  ]
+}
+```
+
+Use a model ID that appears in your local Free-Way model catalog. If gateway auth is disabled, the API key can be any non-empty placeholder value.
+
+## 3. Verify the connection
+
+Run a simple chat request from Continue, then check Free-Way's **Usage** tab to confirm the request reached your local gateway.
+
+## Troubleshooting
+
+| Problem | Check |
+| --- | --- |
+| Connection refused | Free-Way is running on `localhost:8787` |
+| Model not found | The configured model exists in the Free-Way model catalog |
+| Unauthorized | `FREEWAY_API_KEY` matches the gateway key configured in Free-Way |
+| Requests do not appear in Usage | Continue is using this model entry and `apiBase` includes `/v1` |