diff --git a/src/auto-reply/reply/agent-runner-execution.test.ts b/src/auto-reply/reply/agent-runner-execution.test.ts
index 27757c1ffeaf..cbec84628f88 100644
--- a/src/auto-reply/reply/agent-runner-execution.test.ts
+++ b/src/auto-reply/reply/agent-runner-execution.test.ts
@@ -7867,10 +7867,6 @@ describe("runAgentTurnWithFallback", () => {
       reason: "timeout" as const,
       message: "provider request timed out without status token",
     },
-    {
-      reason: "overloaded" as const,
-      message: "provider capacity exhausted briefly",
-    },
   ])(
     "retries once for structured FailoverError $reason without leading HTTP status text",
     async ({ reason, message }) => {
diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index c0fca6f5da1f..217ceeddc299 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -3084,12 +3084,11 @@ async function runAgentTurnWithFallbackInternal(
         !shouldSurfaceToControlUi
           ? classifyProviderRequestError(err)
           : undefined;
+      // Typed overloaded failures stay out of the transient retry: they surface
+      // dedicated overloaded copy immediately instead of being retried silently.
       const isTransientHttp =
         isTransientHttpError(message) ||
-        (isFailoverError(err) &&
-          (err.reason === "timeout" ||
-            err.reason === "server_error" ||
-            err.reason === "overloaded"));
+        (isFailoverError(err) && (err.reason === "timeout" || err.reason === "server_error"));
 
       // Drain/restart aborts stay silent and defer to post-restart
       // main-session recovery, which resumes the interrupted turn (or emits its