diff --git a/agent/src/api/chat.ts b/agent/src/api/chat.ts
index f5151af2..9b227780 100644
--- a/agent/src/api/chat.ts
+++ b/agent/src/api/chat.ts
@@ -38,7 +38,7 @@ export const myProvider = customProvider({
     'sonar-reasoning': perplexity('sonar-reasoning'),
     'gpt-4o': openai('gpt-4o'),
     'gpt-4o-mini': openai('gpt-4o-mini'),
-    'gpt-4.1': openai('gpt-4.1'),
+    'gpt-4.1': openai.responses('gpt-4.1'),
     'gpt-4.1-mini': openai('gpt-4.1-mini'),
     'gpt-4.1-nano': openai('gpt-4.1-nano'),
     'o3-mini': openai('o3-mini'),
diff --git a/agent/src/server.ts b/agent/src/server.ts
index 22a7d812..5c242dee 100644
--- a/agent/src/server.ts
+++ b/agent/src/server.ts
@@ -596,12 +596,25 @@ fastify.route<{ Body: UserChatStreamInput }>({
           }),
         },
         maxSteps: 5,
+        maxRetries: 0,
+        experimental_providerMetadata: {
+          openai: {
+            previousResponseId:
+              'resp_680be1f1906c8192a31f34bdaddd5673065d082b4685d13f',
+          },
+        },
         experimental_generateMessageId: crypto.randomUUID,
         experimental_telemetry: { isEnabled: true, functionId: 'stream-text' },
         onError(error) {
           log.error({ error }, 'Error in chat stream');
         },
         async onFinish(event) {
+          if (event?.providerMetadata?.openai?.responseId) {
+            stream.appendMessageAnnotation({
+              responseId: event.providerMetadata.openai.responseId,
+            });
+          }
+
           if (event.sources.length === 0) {
             await stream.close();
             log.info({}, 'stream closed');