diff --git a/apps/server/src/app.ts b/apps/server/src/app.ts
index 3c5f4bdb72..bb9097ba63 100644
--- a/apps/server/src/app.ts
+++ b/apps/server/src/app.ts
@@ -55,7 +55,16 @@ export default async function buildApp() {
     });
 
     if (!utils.isElectron) {
-        app.use(compression()); // HTTP compression
+        app.use(compression({
+            // Skip compression for SSE endpoints to enable real-time streaming
+            filter: (req, res) => {
+                // Skip compression for LLM chat streaming endpoint
+                if (req.path === "/api/llm-chat/stream") {
+                    return false;
+                }
+                return compression.filter(req, res);
+            }
+        }));
     }
 
     let resourcePolicy = config["Network"]["corsResourcePolicy"] as 'same-origin' | 'same-site' | 'cross-origin' | undefined;
diff --git a/apps/server/src/routes/api/llm_chat.ts b/apps/server/src/routes/api/llm_chat.ts
index b8293d02e0..672df99a4c 100644
--- a/apps/server/src/routes/api/llm_chat.ts
+++ b/apps/server/src/routes/api/llm_chat.ts
@@ -25,21 +25,29 @@ async function streamChat(req: Request, res: Response) {
         return;
     }
 
-    // Set up SSE headers
+    // Set up SSE headers - disable compression and buffering for real-time streaming
     res.setHeader("Content-Type", "text/event-stream");
-    res.setHeader("Cache-Control", "no-cache");
+    res.setHeader("Cache-Control", "no-cache, no-transform");
     res.setHeader("Connection", "keep-alive");
     res.setHeader("X-Accel-Buffering", "no"); // Disable nginx buffering
+    res.setHeader("Content-Encoding", "none"); // Disable compression
     res.flushHeaders();
 
     // Mark response as handled to prevent double-handling by apiResultHandler
     (res as any).triliumResponseHandled = true;
 
+    // Type assertion for flush method (available when compression is used)
+    const flushableRes = res as Response & { flush?: () => void };
+
     try {
         const provider = getProvider(config.provider || "anthropic");
 
         for await (const chunk of provider.streamCompletion(messages, config)) {
             res.write(`data: ${JSON.stringify(chunk)}\n\n`);
+            // Flush immediately to ensure real-time streaming
+            if (typeof flushableRes.flush === "function") {
+                flushableRes.flush();
+            }
         }
     } catch (error) {
         const errorMessage = error instanceof Error ? error.message : "Unknown error";