feat: add safeguards for remote media URLs (#34892)

docs/docs/advanced_features/server_arguments.mdx

@@ -56,6 +56,15 @@ You can find all arguments by `python3 -m sglang.launch_server --help`
 - To enable fp8 weight quantization, add `--quantization fp8` on a fp16 checkpoint or directly load a fp8 checkpoint without specifying any arguments.
 - To enable fp8 kv cache quantization, add `--kv-cache-dtype fp8_e4m3` or `--kv-cache-dtype fp8_e5m2`.
 - To enable deterministic inference and batch invariant operations, add `--enable-deterministic-inference`. More details can be found in [deterministic inference document](./deterministic_inference).
+- If a multimodal server accepts requests from untrusted clients, restrict remote image, video, and audio URLs with `--allowed-media-domains`. SGLang checks the initial URL and every redirect destination against the exact-hostname allowlist. Remote media downloads are limited to 64 MiB by default; adjust `--media-url-max-file-size-mb` when larger trusted media is required.
+
+  ```bash Command
+  python -m sglang.launch_server \
+    --model-path Qwen/Qwen2.5-VL-7B-Instruct \
+    --allowed-media-domains upload.wikimedia.org raw.githubusercontent.com
+  ```
+
+  Without `--allowed-media-domains`, HTTP(S) media from any domain remains allowed for backward compatibility. Do not expose that configuration to untrusted users. Local paths and `data:` URLs are not governed by the domain allowlist.
 - To enable decode context parallelism for MLA models, add `--dcp-size N`. See [Decode Context Parallelism](./dcp).
 - If the model does not have a chat template in the Hugging Face tokenizer, you can specify a [custom chat template](../references/custom_chat_template). If the tokenizer has multiple named templates (e.g., 'default', 'tool_use'), you can select one using `--hf-chat-template-name tool_use`.
 - To run tensor parallelism on multiple nodes, add `--nnodes 2`. If you have two nodes with two GPUs on each node and want to run TP=4, let `sgl-dev-0` be the hostname of the first node and `50000` be an available port, you can use the following commands. If you meet deadlock, please try to add `--disable-cuda-graph`
@@ -3293,6 +3302,18 @@ Please consult the documentation below and [server_args.py](https://github.com/s
       <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.02)"}}><code>&#123;&#125;</code></td>
       <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Type: JSON / Dict</td>
     </tr>
+    <tr>
+      <td style={{padding: "9px 12px", fontWeight: 500, backgroundColor: "rgba(255,255,255,0.02)"}}>`--allowed-media-domains`</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Restrict client-supplied HTTP(S) media URLs and redirect destinations to these exact hostnames.</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.02)"}}>Unrestricted</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Space-separated hostnames</td>
+    </tr>
+    <tr>
+      <td style={{padding: "9px 12px", fontWeight: 500, backgroundColor: "rgba(255,255,255,0.02)"}}>`--media-url-max-file-size-mb`</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Maximum streamed size in MiB for one remote media download. Set to 0 to disable the limit.</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.02)"}}>`64`</td>
+      <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Type: int</td>
+    </tr>
     <tr>
       <td style={{padding: "9px 12px", fontWeight: 500, backgroundColor: "rgba(255,255,255,0.02)"}}>`--mm-enable-dp-encoder`</td>
       <td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Enabling data parallelism for mm encoder. The dp size will be set to the tp size automatically.</td>