diff --git a/src/gateway/protocol/schema.ts b/src/gateway/protocol/schema.ts
index b2c5f8eeb..ffd1c025d 100644
--- a/src/gateway/protocol/schema.ts
+++ b/src/gateway/protocol/schema.ts
@@ -458,7 +458,7 @@ export const CronRunLogEntrySchema = Type.Object(
 export const ChatHistoryParamsSchema = Type.Object(
   {
     sessionKey: NonEmptyString,
-    limit: Type.Optional(Type.Integer({ minimum: 1 })),
+    limit: Type.Optional(Type.Integer({ minimum: 1, maximum: 1000 })),
   },
   { additionalProperties: false },
 );
diff --git a/src/gateway/server.test.ts b/src/gateway/server.test.ts
index cbc1705bb..3992165b1 100644
--- a/src/gateway/server.test.ts
+++ b/src/gateway/server.test.ts
@@ -1786,8 +1786,8 @@ describe("gateway server", () => {
     );
     expect(defaultRes.ok).toBe(true);
     const defaultMsgs = defaultRes.payload?.messages ?? [];
-    expect(defaultMsgs.length).toBe(300);
-    expect(firstContentText(defaultMsgs[0])).toBe("m0");
+    expect(defaultMsgs.length).toBe(200);
+    expect(firstContentText(defaultMsgs[0])).toBe("m100");
 
     const limitedRes = await rpcReq<{ messages?: unknown[] }>(
       ws,
@@ -1829,21 +1829,21 @@ describe("gateway server", () => {
     );
     expect(cappedRes.ok).toBe(true);
     const cappedMsgs = cappedRes.payload?.messages ?? [];
-    expect(cappedMsgs.length).toBe(1000);
-    expect(firstContentText(cappedMsgs[0])).toBe("b500");
+    expect(cappedMsgs.length).toBe(200);
+    expect(firstContentText(cappedMsgs[0])).toBe("b1300");
 
-    const overLimitRes = await rpcReq<{ messages?: unknown[] }>(
+    const maxRes = await rpcReq<{ messages?: unknown[] }>(
       ws,
       "chat.history",
       {
         sessionKey: "main",
-        limit: 5000,
+        limit: 1000,
       },
     );
-    expect(overLimitRes.ok).toBe(true);
-    const overLimitMsgs = overLimitRes.payload?.messages ?? [];
-    expect(overLimitMsgs.length).toBe(1000);
-    expect(firstContentText(overLimitMsgs[0])).toBe("b500");
+    expect(maxRes.ok).toBe(true);
+    const maxMsgs = maxRes.payload?.messages ?? [];
+    expect(maxMsgs.length).toBe(1000);
+    expect(firstContentText(maxMsgs[0])).toBe("b500");
 
     ws.close();
     await server.close();
diff --git a/src/gateway/server.ts b/src/gateway/server.ts
index 28d2a8b50..fe648b737 100644
--- a/src/gateway/server.ts
+++ b/src/gateway/server.ts
@@ -1824,7 +1824,8 @@ export async function startGatewayServer(
                 ? readSessionMessages(sessionId, storePath)
                 : [];
             const hardMax = 1000;
-            const requested = typeof limit === "number" ? limit : hardMax;
+            const defaultLimit = 200;
+            const requested = typeof limit === "number" ? limit : defaultLimit;
             const max = Math.min(hardMax, requested);
             const messages =
               rawMessages.length > max ? rawMessages.slice(-max) : rawMessages;