diff --git a/docs/plugins/google-meet.md b/docs/plugins/google-meet.md
index 3a8b1720d6e..3bc41cbcd10 100644
--- a/docs/plugins/google-meet.md
+++ b/docs/plugins/google-meet.md
@@ -12,6 +12,9 @@ Google Meet participant support for OpenClaw — the plugin is explicit by desig
 - `realtime` voice is the default mode.
 - Realtime voice can call back into the full OpenClaw agent when deeper
   reasoning or tools are needed.
+- Agents choose the join behavior with `mode`: use `realtime` for live
+  listen/talk-back, or `transcribe` to join/control the browser without the
+  realtime voice bridge.
 - Auth starts as personal Google OAuth or an already signed-in Chrome profile.
 - There is no automatic consent announcement.
 - The default Chrome audio backend is `BlackHole 2ch`.
@@ -79,10 +82,16 @@ Or let an agent join through the `google_meet` tool:
 ```json
 {
   "action": "join",
-  "url": "https://meet.google.com/abc-defg-hij"
+  "url": "https://meet.google.com/abc-defg-hij",
+  "transport": "chrome-node",
+  "mode": "realtime"
 }
 ```
 
+For an observe-only/browser-control join, set `"mode": "transcribe"`. That does
+not start the duplex realtime model bridge, so it will not talk back into the
+meeting.
+
 Chrome joins as the signed-in Chrome profile. In Meet, pick `BlackHole 2ch` for
 the microphone/speaker path used by OpenClaw. For clean duplex audio, use
 separate virtual devices or a Loopback-style graph; a single BlackHole device is
diff --git a/extensions/google-meet/index.ts b/extensions/google-meet/index.ts
index 9b5b13c147b..c9b64993728 100644
--- a/extensions/google-meet/index.ts
+++ b/extensions/google-meet/index.ts
@@ -35,7 +35,7 @@ const googleMeetConfigSchema = {
     },
     defaultMode: {
       label: "Default Mode",
-      help: "Realtime voice is the default.",
+      help: "Realtime starts the duplex voice model loop. Transcribe joins/observes without the realtime talk-back bridge.",
     },
     "chrome.audioBackend": {
       label: "Chrome Audio Backend",
@@ -148,7 +148,13 @@ const GoogleMeetToolSchema = Type.Object({
   transport: Type.Optional(
     Type.String({ enum: ["chrome", "chrome-node", "twilio"], description: "Join transport" }),
   ),
-  mode: Type.Optional(Type.String({ enum: ["realtime", "transcribe"], description: "Join mode" })),
+  mode: Type.Optional(
+    Type.String({
+      enum: ["realtime", "transcribe"],
+      description:
+        "Join mode. realtime starts live listen/talk-back through the realtime voice model; transcribe joins without the realtime talk-back bridge.",
+    }),
+  ),
   dialInNumber: Type.Optional(Type.String({ description: "Meet dial-in number for Twilio" })),
   pin: Type.Optional(Type.String({ description: "Meet phone PIN for Twilio" })),
   dtmfSequence: Type.Optional(Type.String({ description: "Explicit DTMF sequence for Twilio" })),
diff --git a/extensions/google-meet/src/cli.ts b/extensions/google-meet/src/cli.ts
index 8001330b07e..8dff6026db7 100644
--- a/extensions/google-meet/src/cli.ts
+++ b/extensions/google-meet/src/cli.ts
@@ -177,7 +177,10 @@ export function registerGoogleMeetCli(params: {
     .command("join")
     .argument("[url]", "Explicit https://meet.google.com/... URL")
     .option("--transport <transport>", "Transport: chrome, chrome-node, or twilio")
-    .option("--mode <mode>", "Mode: realtime or transcribe")
+    .option(
+      "--mode <mode>",
+      "Mode: realtime for live talk-back, transcribe to join without the realtime voice bridge",
+    )
     .option("--message <text>", "Realtime speech to trigger after join")
     .option("--dial-in-number <phone>", "Meet dial-in number for Twilio transport")
     .option("--pin <pin>", "Meet phone PIN; # is appended if omitted")
@@ -200,7 +203,10 @@ export function registerGoogleMeetCli(params: {
     .command("test-speech")
     .argument("[url]", "Explicit https://meet.google.com/... URL")
     .option("--transport <transport>", "Transport: chrome, chrome-node, or twilio")
-    .option("--mode <mode>", "Mode: realtime or transcribe")
+    .option(
+      "--mode <mode>",
+      "Mode: realtime for live talk-back, transcribe to join without the realtime voice bridge",
+    )
     .option(
       "--message <text>",
       "Realtime speech to trigger",