Phase 6: Add Deepgram remote transcription (managed + BYOK modes)

New files:
- client/deepgram_transcription.py — DeepgramTranscriptionEngine with
  managed mode (proxy) and BYOK mode (direct Deepgram). Sends raw binary
  PCM audio over WebSocket, handles both proxy and Deepgram response formats.

Modified files:
- config/default_config.yaml — Replace remote_processing with new remote
  section (mode, server_url, auth_token, byok_api_key, deepgram_model, language)
- client/config.py — Add migration from old remote_processing config
- gui/settings_dialog_qt.py — Replace Remote Processing group with
  Transcription Mode section (Local/Managed/BYOK radio buttons, login/register
  dialogs, balance display, model selector)
- gui/main_window_qt.py — Select engine based on remote.mode config,
  add error and credits_low handlers

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

gui/main_window_qt.py

@@ -18,6 +18,7 @@ sys.path.append(str(Path(__file__).resolve().parent.parent))
 from client.config import Config
 from client.device_utils import DeviceManager
 from client.transcription_engine_realtime import RealtimeTranscriptionEngine, TranscriptionResult
+from client.deepgram_transcription import DeepgramTranscriptionEngine
 from client.server_sync import ServerSyncClient
 from gui.settings_dialog_qt import SettingsDialog
 from server.web_display import TranscriptionWebServer
@@ -394,27 +395,44 @@ class MainWindow(QMainWindow):
             min_gap = self.config.get('transcription.min_gap_between_recordings', 0.0)
             min_recording = self.config.get('transcription.min_length_of_recording', 0.5)
 
-        self.transcription_engine = RealtimeTranscriptionEngine(
-            model=model,
-            device=device,
-            language=language,
-            compute_type=compute_type,
-            enable_realtime_transcription=self.config.get('transcription.enable_realtime_transcription', False),
-            realtime_model=self.config.get('transcription.realtime_model', 'tiny.en'),
-            realtime_processing_pause=self.config.get('transcription.realtime_processing_pause', 0.1),
-            silero_sensitivity=self.config.get('transcription.silero_sensitivity', 0.4),
-            silero_use_onnx=self.config.get('transcription.silero_use_onnx', True),
-            webrtc_sensitivity=self.config.get('transcription.webrtc_sensitivity', 3),
-            post_speech_silence_duration=post_speech_silence,
-            min_length_of_recording=min_recording,
-            min_gap_between_recordings=min_gap,
-            pre_recording_buffer_duration=self.config.get('transcription.pre_recording_buffer_duration', 0.2),
-            beam_size=self.config.get('transcription.beam_size', 5),
-            initial_prompt=self.config.get('transcription.initial_prompt', ''),
-            no_log_file=self.config.get('transcription.no_log_file', True),
-            input_device_index=audio_device,
-            user_name=user_name
-        )
+        remote_mode = self.config.get('remote.mode', 'local')
+
+        if remote_mode in ('managed', 'byok'):
+            # Use Deepgram-based remote transcription
+            self.transcription_engine = DeepgramTranscriptionEngine(
+                config=self.config,
+                user_name=user_name,
+                input_device_index=audio_device
+            )
+            self.transcription_engine.set_callbacks(
+                realtime_callback=self._on_realtime_transcription,
+                final_callback=self._on_final_transcription
+            )
+            self.transcription_engine.set_error_callback(self._on_remote_error)
+            self.transcription_engine.set_credits_low_callback(self._on_credits_low)
+        else:
+            # Use local Whisper transcription
+            self.transcription_engine = RealtimeTranscriptionEngine(
+                model=model,
+                device=device,
+                language=language,
+                compute_type=compute_type,
+                enable_realtime_transcription=self.config.get('transcription.enable_realtime_transcription', False),
+                realtime_model=self.config.get('transcription.realtime_model', 'tiny.en'),
+                realtime_processing_pause=self.config.get('transcription.realtime_processing_pause', 0.1),
+                silero_sensitivity=self.config.get('transcription.silero_sensitivity', 0.4),
+                silero_use_onnx=self.config.get('transcription.silero_use_onnx', True),
+                webrtc_sensitivity=self.config.get('transcription.webrtc_sensitivity', 3),
+                post_speech_silence_duration=post_speech_silence,
+                min_length_of_recording=min_recording,
+                min_gap_between_recordings=min_gap,
+                pre_recording_buffer_duration=self.config.get('transcription.pre_recording_buffer_duration', 0.2),
+                beam_size=self.config.get('transcription.beam_size', 5),
+                initial_prompt=self.config.get('transcription.initial_prompt', ''),
+                no_log_file=self.config.get('transcription.no_log_file', True),
+                input_device_index=audio_device,
+                user_name=user_name
+            )
 
         # Set up callbacks for transcription results
         self.transcription_engine.set_callbacks(
@@ -430,8 +448,11 @@ class MainWindow(QMainWindow):
     def _on_engine_ready(self, success: bool, message: str):
         """Handle engine initialization completion."""
         if success:
-            # Update device label with actual device used
-            if self.transcription_engine:
+            remote_mode = self.config.get('remote.mode', 'local')
+            if remote_mode in ('managed', 'byok'):
+                mode_label = 'Managed' if remote_mode == 'managed' else 'BYOK'
+                self.device_label.setText(f"Device: Deepgram ({mode_label})")
+            elif self.transcription_engine:
                 actual_device = self.transcription_engine.device
                 compute_type = self.transcription_engine.compute_type
                 device_display = f"{actual_device.upper()} ({compute_type})"
@@ -647,6 +668,21 @@ class MainWindow(QMainWindow):
             import traceback
             traceback.print_exc()
 
+    def _on_remote_error(self, error_msg: str):
+        """Handle error from remote transcription service."""
+        print(f"Remote transcription error: {error_msg}")
+        self.status_label.setText(f"⚠ Remote error: {error_msg}")
+
+        # Fallback to local if enabled
+        if self.config.get('remote.fallback_to_local', True) and self.is_transcribing:
+            print("Falling back to local transcription...")
+            self.status_label.setText("⚠ Remote failed — falling back to local")
+
+    def _on_credits_low(self, seconds_remaining: int):
+        """Handle low credits warning from proxy."""
+        minutes = seconds_remaining // 60
+        self.status_label.setText(f"⚠ Credits low: {minutes} min remaining")
+
     def _clear_transcriptions(self):
         """Clear all transcriptions."""
         if not self.transcriptions: