Add LFM to recommended local models (#24812)

* Fix audio worker startup after enabling a disabled camera

* Document audio lifecycle test helpers

* Focus audio startup coverage on the enabled update regression

docs/docs/configuration/genai/config.md

@@ -75,11 +75,12 @@ Running Generative AI models on CPU is not recommended, as high inference times
 
 You must use a vision-capable model with Frigate. The following models are recommended for local deployment of the `descriptions` and `chat` roles:
 
-| Model               | Review [frame mode](/configuration/genai/genai_review#frame-mode) | Notes                                                                                                                                                                |
-| ------------------- | --------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `qwen3-vl`          | `frames`                                                              | Strong visual and situational understanding, enhanced ability to identify smaller objects and interactions with object. Follows a sequence of frames on its own.     |
-| `qwen3.6`/`qwen3.8` | `frames`                                                              | Strong situational understanding, but missing DeepStack from qwen3-vl leading to worse performance for identifying objects in people's hand and other small details. |
-| `gemma4`            | `annotated_frames`                                                    | Strong situational understanding, sometimes resorts to more vague terms like 'interacts' instead of assigning a specific action. Loses track of activity that repeats or reverses, so it benefits from annotated frames. |
+| Model               | Review [frame mode](/configuration/genai/genai_review#frame-mode) | Review [response style](/configuration/genai/genai_review#response-style) | Notes                                                                                                                                                                                                                    |
+| ------------------- | ----------------------------------------------------------------- | ------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
+| `qwen3-vl`          | `frames`                                                          | `natural`                                                                 | Strong visual and situational understanding, enhanced ability to identify smaller objects and interactions with object. Follows a sequence of frames on its own.                                                         |
+| `qwen3.6`/`qwen3.8` | `frames`                                                          | `natural`                                                                 | Strong situational understanding, but missing DeepStack from qwen3-vl leading to worse performance for identifying objects in people's hand and other small details.                                                     |
+| `gemma4`            | `annotated_frames`                                                | `default`                                                                 | Strong situational understanding, sometimes resorts to more vague terms like 'interacts' instead of assigning a specific action. Loses track of activity that repeats or reverses, so it benefits from annotated frames. |
+| `lfm3-vl`           | `annotated_frames`                                                | `concise`                                                                 | Very efficient with fast inference, making it a good choice for less powerful hardware or high volumes of review items.                                                                                                  |
 
 #### Embedding models
 
@@ -95,7 +96,7 @@ The `embeddings` role needs a different kind of model. Text queries are matched
 The `transcribe` role needs a model that accepts audio input. A text-only or vision-only model cannot serve this role. The following are recommended for local deployment of the `transcribe` role:
 
 | Model       | Notes                                                                                                                                                                                                      |
-| ----------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
+| ----------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
 | `qwen3-asr` | Dedicated speech recognition model covering 30 languages, and the better choice for transcription quality. It only transcribes, so it cannot be shared with the `descriptions` or `chat` roles.            |
 | `gemma4`    | General multimodal model that accepts audio as well as images, so one served model can cover `transcribe` alongside the other roles. Transcript quality is below `qwen3-asr`, particularly on noisy audio. |
 

frigate/events/audio.py

@@ -178,6 +178,7 @@ class AudioProcessor(FrigateProcess):
             [
                 CameraConfigUpdateEnum.add,
                 CameraConfigUpdateEnum.audio,
+                CameraConfigUpdateEnum.enabled,
                 CameraConfigUpdateEnum.ffmpeg,
                 CameraConfigUpdateEnum.remove,
             ],

frigate/test/test_audio_spawn.py

@@ -1,4 +1,4 @@
-"""Tests for audio maintainer spawning on runtime camera adds."""
+"""Tests for audio maintainer spawning on runtime camera updates."""
 
 import threading
 import unittest
@@ -42,6 +42,42 @@ class TestAudioSpawnIfNeeded(unittest.TestCase):
         processor.logger = MagicMock()
         return processor
 
+    def test_initially_disabled_camera_spawns_after_enabled_update(self):
+        """Consume an enabled update through the manager's real subscriber."""
+        config = _build_config()
+        config.cameras["front_door"].enabled = False
+        stop = MagicMock()
+        stop.wait.side_effect = [False, True]
+        processor = AudioProcessor(
+            config, {"front_door": MagicMock()}, MagicMock(), stop
+        )
+        processor.logger = MagicMock()
+        updates = iter([("config/cameras/front_door/enabled", True), (None, None)])
+        self.addCleanup(
+            setattr, threading.current_thread(), "name", threading.current_thread().name
+        )
+
+        with (
+            patch.object(processor, "pre_run_setup"),
+            patch("frigate.config.camera.updater.ConfigSubscriber") as transport,
+            patch("frigate.events.audio.AudioEventMaintainer") as maintainer,
+        ):
+
+            def receive_update():
+                self.assertEqual(processor.audio_threads, {})
+                maintainer.assert_not_called()
+                return next(updates)
+
+            transport.return_value.check_for_update.side_effect = receive_update
+            maintainer.return_value.is_alive.return_value = False
+            processor.run()
+
+        self.assertEqual(
+            processor.audio_threads, {"front_door": maintainer.return_value}
+        )
+        maintainer.assert_called_once()
+        maintainer.return_value.start.assert_called_once()
+
     def test_skips_camera_whose_metrics_have_not_been_created_yet(self):
         """The camera maintainer creates metrics on its own poll of the add
         update, so the audio processor can see the camera first."""