Update app.py

app.py

@@ -58,10 +58,12 @@ class GeminiHandler(AsyncAudioVideoStreamHandler):
         return GeminiHandler()
 
     async def start_up(self):
+        await self.wait_for_args()
+        api_key = self.latest_args[3]
         client = genai.Client(
-            api_key=os.getenv("GEMINI_API_KEY"), http_options={"api_version": "v1alpha"}
+            api_key=api_key, http_options={"api_version": "v1alpha"}
         )
-        config = {"response_modalities": ["AUDIO"]}
+        config = {"response_modalities": ["AUDIO"], "system_instruction": "You are an art history teacher that will describe the artwork passed in as an image to the user. Describe the history and significance of the artwork."}
         async with client.aio.live.connect(
             model="gemini-2.0-flash-exp",
             config=config,  # type: ignore
@@ -87,8 +89,8 @@ class GeminiHandler(AsyncAudioVideoStreamHandler):
             if time.time() - self.last_frame_time > 1:
                 self.last_frame_time = time.time()
                 await self.session.send(input=encode_image(frame))
-                if self.latest_args[1] is not None:
-                    await self.session.send(input=encode_image(self.latest_args[1]))
+                if self.latest_args[2] is not None:
+                    await self.session.send(input=encode_image(self.latest_args[2]))
 
     async def video_emit(self):
         frame = await wait_for_item(self.video_queue, 0.01)
@@ -121,10 +123,16 @@ stream = Stream(
     handler=GeminiHandler(),
     modality="audio-video",
     mode="send-receive",
-    rtc_configuration=get_cloudflare_turn_credentials_async,
+    rtc_configuration=get_cloudflare_turn_credentials_async if get_space() else None,
     time_limit=180 if get_space() else None,
     additional_inputs=[
-        gr.Image(label="Image", type="numpy", sources=["upload", "clipboard"])
+        gr.Markdown(
+            "## 🎨 Art History Teacher\n\n"
+            "Provide an image of the artwork and Gemini will describe it to you."
+            "To get a Gemini API key, please visit the [Gemini API Key](https://console.cloud.google.com/apis/api/generativelanguage.googleapis.com/credentials) page."
+        ),
+        gr.Image(label="Image", type="numpy", sources=["upload", "clipboard"]),
+        gr.Textbox(label="Gemini API Key"),
     ],
     ui_args={
         "icon": "https://www.gstatic.com/lamda/images/gemini_favicon_f069958c85030456e93de685481c559f160ea06b.png",
@@ -135,9 +143,4 @@ stream = Stream(
 )
 
 if __name__ == "__main__":
-    if (mode := os.getenv("MODE")) == "UI":
-        stream.ui.launch(server_port=7860)
-    elif mode == "PHONE":
-        raise ValueError("Phone mode not supported for this demo")
-    else:
-        stream.ui.launch(server_port=7860)
+    stream.ui.launch()