Merge 815a2017c2 into d556ff8df2

* add review padding to explore debug replay api calls

* add semantic search model size widget

disables model_size select with n/a text when an embeddings genai provider is selected

* regenerate zone contours and per-zone filter masks on detect resolution change

* treat null as a clear sentinel in buildOverrides so nullable field edits don't snap back

* extract replay config sheet to new component

* add validation and messages for detect settings

docker/tensorrt/requirements-amd64.txt

@@ -15,4 +15,4 @@ nvidia-nccl-cu12==2.26.2.post1; platform_machine == 'x86_64'
 nvidia-nvjitlink-cu12==12.8.93; platform_machine == 'x86_64'
 onnx==1.16.*; platform_machine == 'x86_64'
 onnxruntime-gpu==1.24.*; platform_machine == 'x86_64'
-protobuf==3.20.3; platform_machine == 'x86_64'
+protobuf==7.35.0; platform_machine == 'x86_64'

docker/tensorrt/requirements-models-arm64.txt

@@ -1,2 +1,2 @@
 onnx == 1.14.0; platform_machine == 'aarch64'
-protobuf == 3.20.3; platform_machine == 'aarch64'
+protobuf == 7.35.0; platform_machine == 'aarch64'

docs/docs/configuration/genai/config.md

@@ -49,15 +49,14 @@ You should have at least 8 GB of RAM available (or VRAM if running on GPU) to ru
 
 ### Model Types: Instruct vs Thinking
 
-Most vision-language models are available as **instruct** models, which are fine-tuned to follow instructions and respond concisely to prompts. However, some models (such as certain Qwen-VL or minigpt variants) offer both **instruct** and **thinking** versions.
+Vision-language models come in **instruct** variants (fine-tuned to follow instructions and respond concisely), **thinking** variants (fine-tuned for free-form, speculative reasoning), and **hybrid** variants that support both modes per request. Most modern vision-language models are hybrid.
 
-- **Instruct models** are always recommended for use with Frigate. These models generate direct, relevant, actionable descriptions that best fit Frigate's object and event summary use case.
-- **Reasoning / Thinking models** are fine-tuned for more free-form, open-ended, and speculative outputs, which are typically not concise and may not provide the practical summaries Frigate expects. For this reason, Frigate does **not** recommend or support using thinking models.
+Frigate manages reasoning per task automatically:
 
-Some models are labeled as **hybrid** (capable of both thinking and instruct tasks). In these cases, it is recommended to disable reasoning / thinking, which is generally model specific (see your models documentation).
+- **Description tasks** (object descriptions, review descriptions, review summaries) are synthesis-only and benefit from concise, direct output, so Frigate disables thinking for these calls when the model exposes a per-request toggle.
+- **Chat** lets you toggle thinking on or off from the composer when the configured model supports it.
 
-**Recommendation:**
-Always select the `-instruct` or documented instruct/tagged variant of any model you use in your Frigate configuration. If in doubt, refer to your model provider's documentation or model library for guidance on the correct model variant to use.
+You can use a pure instruct, hybrid, or thinking-capable model with Frigate — no extra configuration is required to disable thinking for descriptions.
 
 ### llama.cpp
 

docs/docs/configuration/review.md

@@ -23,7 +23,7 @@ In 0.14 and later, all of that is bundled into a single review item which starts
 
 ## Alerts and Detections
 
-Not every segment of video captured by Frigate may be of the same level of interest to you. Video of people who enter your property may be a different priority than those walking by on the sidewalk. For this reason, Frigate 0.14 categorizes review items as _alerts_ and _detections_. By default, all person and car objects are considered alerts. You can refine categorization of your review items by configuring required zones for them.
+Not every segment of video captured by Frigate may be of the same level of interest to you. Video of people who enter your property may be a different priority than those walking by on the sidewalk. For this reason, Frigate categorizes review items as _alerts_ and _detections_. By default, all person and car objects are considered alerts. You can refine categorization of your review items by configuring required zones for them.
 
 :::note
 

docs/docs/troubleshooting/dummy-camera.md

@@ -56,6 +56,7 @@ Only one replay session can be active at a time. If a session is already running
 - The replay will not always produce identical results to the original run. Different frames may be selected on replay, which can change detections and tracking.
 - Motion detection depends on the exact frames used; small frame shifts can change motion regions and therefore what gets passed to the detector.
 - Object detection is not fully deterministic: models and post-processing can yield slightly different results across runs.
+- In cases where a detection is short and a replay may only be a small number of frames, it is recommended to manually add some padding before and after the detection so that the motion and object detectors have time to settle into the scene. Rather than starting Debug Replay from Explore, navigate to History for your camera, choose Debug Replay from the Actions menu, and click the "From Timeline" or "Custom" option.
 
 Treat the replay as a close approximation rather than an exact reproduction. Run multiple loops and examine the debug overlays and logs to understand the behavior.
 

frigate/api/app.py

@@ -750,6 +750,40 @@ def _config_set_in_memory(request: Request, body: AppConfigSetBody) -> JSONRespo
                     settings,
                 )
 
+                # detect resize also republishes motion + objects so other
+                # processes pick up the rebuilt masks, and fires refresh so
+                # the camera maintainer recycles the camera process to pick
+                # up the new ffmpeg cmd / SHM sizing
+                if field == "detect":
+                    cam_cfg = config.cameras.get(camera)
+                    if cam_cfg is not None:
+                        if cam_cfg.motion is not None:
+                            request.app.config_publisher.publish_update(
+                                CameraConfigUpdateTopic(
+                                    CameraConfigUpdateEnum.motion, camera
+                                ),
+                                cam_cfg.motion,
+                            )
+                        request.app.config_publisher.publish_update(
+                            CameraConfigUpdateTopic(
+                                CameraConfigUpdateEnum.objects, camera
+                            ),
+                            cam_cfg.objects,
+                        )
+                        if cam_cfg.zones:
+                            request.app.config_publisher.publish_update(
+                                CameraConfigUpdateTopic(
+                                    CameraConfigUpdateEnum.zones, camera
+                                ),
+                                cam_cfg.zones,
+                            )
+                        request.app.config_publisher.publish_update(
+                            CameraConfigUpdateTopic(
+                                CameraConfigUpdateEnum.refresh, camera
+                            ),
+                            cam_cfg,
+                        )
+
         return JSONResponse(
             content={"success": True, "message": "Config applied in-memory"},
             status_code=200,

frigate/camera/maintainer.py

@@ -14,6 +14,7 @@ from frigate.config.camera.updater import (
     CameraConfigUpdateEnum,
     CameraConfigUpdateSubscriber,
 )
+from frigate.const import REPLAY_CAMERA_PREFIX
 from frigate.models import Regions
 from frigate.util.builtin import empty_and_close_queue
 from frigate.util.image import SharedMemoryFrameManager, UntrackedSharedMemory
@@ -50,6 +51,7 @@ class CameraMaintainer(threading.Thread):
             [
                 CameraConfigUpdateEnum.add,
                 CameraConfigUpdateEnum.remove,
+                CameraConfigUpdateEnum.refresh,
             ],
         )
         self.shm_count = self.__calculate_shm_frame_count()
@@ -202,6 +204,25 @@ class CameraMaintainer(threading.Thread):
                 capture_process.terminate()
                 capture_process.join()
 
+    def __unlink_camera_frame_slots(self, camera: str) -> None:
+        """Drop the camera's per-frame YUV SHM segments from this
+        process's frame_manager and unlink them at the OS level.
+
+        Safe to call after the camera's capture/processor subprocesses
+        have been joined — they no longer hold mappings, so unlink frees
+        the segments immediately. Other long-lived processes that opened
+        these slots will continue using their existing mappings until
+        they call frame_manager.get with a shape that no longer fits
+        (the get path drops and reopens stale refs).
+        """
+        prefix = f"{camera}_frame"
+        names = [n for n in list(self.frame_manager.shm_store) if n.startswith(prefix)]
+        for name in names:
+            try:
+                self.frame_manager.delete(name)
+            except Exception as exc:
+                logger.debug("Could not unlink SHM %s: %s", name, exc)
+
     def __stop_camera_process(self, camera: str) -> None:
         camera_process = self.camera_processes.get(camera)
         if camera_process is not None:
@@ -253,12 +274,45 @@ class CameraMaintainer(threading.Thread):
                     for camera in updated_cameras:
                         self.__stop_camera_capture_process(camera)
                         self.__stop_camera_process(camera)
+                        self.__unlink_camera_frame_slots(camera)
                         self.capture_processes.pop(camera, None)
                         self.camera_processes.pop(camera, None)
                         self.camera_stop_events.pop(camera, None)
                         self.region_grids.pop(camera, None)
                         self.camera_metrics.pop(camera, None)
                         self.ptz_metrics.pop(camera, None)
+                elif update_type == CameraConfigUpdateEnum.refresh.name:
+                    # Recycle replay cameras so detect width/height/fps
+                    # propagate through ffmpeg args, SHM sizing, and the
+                    # region grid. Regular cameras detect change still
+                    # requires a full restart.
+                    for camera in updated_cameras:
+                        if not camera.startswith(REPLAY_CAMERA_PREFIX):
+                            continue
+
+                        new_config = self.update_subscriber.camera_configs.get(camera)
+                        if new_config is None:
+                            # remove arrived in the same batch
+                            continue
+
+                        if (
+                            camera not in self.camera_processes
+                            and camera not in self.capture_processes
+                        ):
+                            continue
+
+                        # rebuild ffmpeg cmds on the shared config so the
+                        # new subprocesses spawn with current args
+                        new_config.recreate_ffmpeg_cmds()
+
+                        self.__stop_camera_capture_process(camera)
+                        self.__stop_camera_process(camera)
+                        self.__unlink_camera_frame_slots(camera)
+                        self.capture_processes.pop(camera, None)
+                        self.camera_processes.pop(camera, None)
+
+                        self.__start_camera_processor(camera, new_config, runtime=True)
+                        self.__start_camera_capture(camera, new_config, runtime=True)
 
         # ensure the capture processes are done
         for camera in self.capture_processes.keys():

frigate/camera/state.py

@@ -45,6 +45,7 @@ class CameraState:
         self.frame_cache: dict[float, dict[str, Any]] = {}
         self.zone_objects: defaultdict[str, list[Any]] = defaultdict(list)
         self._current_frame = np.zeros(self.camera_config.frame_shape_yuv, np.uint8)
+        self._last_frame_shape: tuple[int, int] = self.camera_config.frame_shape_yuv
         self.current_frame_lock = threading.Lock()
         self.current_frame_time = 0.0
         self.motion_boxes: list[tuple[int, int, int, int]] = []
@@ -303,6 +304,42 @@ class CameraState:
     def on(self, event_type: str, callback: Callable[..., Any]) -> None:
         self.callbacks[event_type].append(callback)
 
+    def _discard_stale_resolution_state(
+        self, current_detections: dict[str, dict[str, Any]]
+    ) -> bool:
+        """Drop tracked state when the camera's detect resolution has
+        changed, and signal the caller to skip this batch if it contains
+        out-of-bounds boxes from the pre-recycle detect process.
+
+        Returns True when the batch should be skipped entirely.
+        """
+        # detect resolution changed — drop tracked state so old-grid
+        # boxes don't leak through end-callbacks
+        current_shape = self.camera_config.frame_shape_yuv
+        if current_shape != self._last_frame_shape:
+            logger.debug(
+                f"{self.name}: detect resolution changed {self._last_frame_shape} -> {current_shape}, dropping tracked state"
+            )
+            with self.current_frame_lock:
+                self.tracked_objects.clear()
+                self.motion_boxes = []
+                self.regions = []
+            self._last_frame_shape = current_shape
+
+        # drop in-flight batches from the pre-recycle detect process
+        # whose boxes exceed the current detect resolution
+        detect = self.camera_config.detect
+        if detect.width is not None and detect.height is not None:
+            for obj in current_detections.values():
+                box = obj.get("box")
+                if box and (box[2] > detect.width or box[3] > detect.height):
+                    logger.debug(
+                        f"{self.name}: dropping stale-resolution detection batch (box {box} exceeds {detect.width}x{detect.height})"
+                    )
+                    return True
+
+        return False
+
     def update(
         self,
         frame_name: str,
@@ -311,6 +348,9 @@ class CameraState:
         motion_boxes: list[tuple[int, int, int, int]],
         regions: list[tuple[int, int, int, int]],
     ) -> None:
+        if self._discard_stale_resolution_state(current_detections):
+            return
+
         current_frame = self.frame_manager.get(
             frame_name, self.camera_config.frame_shape_yuv
         )
@@ -332,14 +372,18 @@ class CameraState:
                 current_detections[id],
             )
 
-            # add initial frame to frame cache
-            logger.debug(
-                f"{self.name}: New object, adding {frame_time} to frame cache for {id}"
-            )
-            self.frame_cache[frame_time] = {
-                "frame": np.copy(current_frame),  # type: ignore[arg-type]
-                "object_id": id,
-            }
+            # Skip caching when the frame buffer isn't readable — e.g.
+            # frame_manager.get returned None because the SHM segment was
+            # unlinked or hasn't been recreated yet during a camera
+            # add/remove cycle.
+            if current_frame is not None:
+                logger.debug(
+                    f"{self.name}: New object, adding {frame_time} to frame cache for {id}"
+                )
+                self.frame_cache[frame_time] = {
+                    "frame": np.copy(current_frame),
+                    "object_id": id,
+                }
 
             # save initial thumbnail data and best object
             thumbnail_data = {

frigate/config/camera/updater.py

@@ -26,6 +26,7 @@ class CameraConfigUpdateEnum(str, Enum):
     object_genai = "object_genai"
     onvif = "onvif"
     record = "record"
+    refresh = "refresh"  # signals the camera maintainer to recycle the camera process
     remove = "remove"  # for removing a camera
     review = "review"
     review_genai = "review_genai"
@@ -84,8 +85,8 @@ class CameraConfigUpdateSubscriber:
         self, camera: str, update_type: CameraConfigUpdateEnum, updated_config: Any
     ) -> None:
         if update_type == CameraConfigUpdateEnum.add:
-            self.config.cameras[camera] = updated_config
-            self.camera_configs[camera] = updated_config
+            shared = self.config.cameras.setdefault(camera, updated_config)
+            self.camera_configs[camera] = shared
             return
         elif update_type == CameraConfigUpdateEnum.remove:
             self.config.cameras.pop(camera, None)

frigate/debug_replay.py

@@ -238,6 +238,10 @@ class DebugReplayManager:
             zone_dump.setdefault("coordinates", zone_config.coordinates)
             zones_dict[zone_name] = zone_dump
 
+        # Extract LPR and face recognition configs
+        lpr_dict = source_config.lpr.model_dump()
+        face_recognition_dict = source_config.face_recognition.model_dump()
+
         # Extract motion config (exclude runtime fields)
         motion_dict = {}
         if source_config.motion is not None:
@@ -287,8 +291,8 @@ class DebugReplayManager:
             },
             "birdseye": {"enabled": False},
             "audio": {"enabled": False},
-            "lpr": {"enabled": False},
-            "face_recognition": {"enabled": False},
+            "lpr": lpr_dict,
+            "face_recognition": face_recognition_dict,
         }
 
     def _cleanup_db(self, camera_name: str) -> None:

frigate/genai/__init__.py

@@ -5,7 +5,7 @@ import json
 import logging
 import os
 import re
-from typing import Any, Callable, Optional
+from typing import Any, AsyncGenerator, Callable, Optional
 
 import numpy as np
 from pydantic import ValidationError
@@ -359,6 +359,41 @@ class GenAIClient:
             "finish_reason": "error",
         }
 
+    async def chat_with_tools_stream(
+        self,
+        messages: list[dict[str, Any]],
+        tools: Optional[list[dict[str, Any]]] = None,
+        tool_choice: Optional[str] = "auto",
+        enable_thinking: Optional[bool] = None,
+    ) -> AsyncGenerator[tuple[str, Any], None]:
+        """Streaming counterpart to `chat_with_tools`.
+
+        Yields ``(kind, value)`` tuples where ``kind`` is one of:
+            - 'content_delta': value is a string fragment of the answer
+            - 'reasoning_delta': value is a string fragment of the reasoning
+              trace (emitted before content for thinking models)
+            - 'stats': value is a usage stats dict
+            - 'message': value is the final dict shape described in
+              `chat_with_tools`
+
+        Argument semantics — including ``enable_thinking`` — match
+        `chat_with_tools`. Providers that don't support streaming should
+        override this and yield an error 'message' event.
+        """
+        logger.warning(
+            f"{self.__class__.__name__} does not support chat_with_tools_stream. "
+            "This method should be overridden by the provider implementation."
+        )
+        yield (
+            "message",
+            {
+                "content": None,
+                "reasoning": None,
+                "tool_calls": None,
+                "finish_reason": "error",
+            },
+        )
+
 
 def load_providers() -> None:
     plugins_dir = os.path.join(os.path.dirname(__file__), "plugins")

frigate/genai/plugins/gemini.py

@@ -1,5 +1,7 @@
 """Gemini Provider for Frigate AI."""
 
+import base64
+import binascii
 import json
 import logging
 from typing import Any, AsyncGenerator, Optional
@@ -14,6 +16,27 @@ from frigate.genai import GenAIClient, register_genai_provider
 logger = logging.getLogger(__name__)
 
 
+def _decode_thought_signature(value: Any) -> Optional[bytes]:
+    """Decode a base64-encoded thought_signature carried across conversation turns."""
+    if not value:
+        return None
+    if isinstance(value, bytes):
+        return value
+    if isinstance(value, str):
+        try:
+            return base64.b64decode(value)
+        except (binascii.Error, ValueError):
+            return None
+    return None
+
+
+def _encode_thought_signature(signature: Optional[bytes]) -> Optional[str]:
+    """Encode bytes thought_signature as base64 so it survives JSON-friendly transport."""
+    if not signature:
+        return None
+    return base64.b64encode(signature).decode("ascii")
+
+
 def _stats_from_gemini_usage(usage: Any) -> Optional[dict[str, Any]]:
     """Build a stats dict from a Gemini usage_metadata object."""
     prompt_tokens = getattr(usage, "prompt_token_count", None)
@@ -169,11 +192,17 @@ class GeminiClient(GenAIClient):
                         if not isinstance(tc_args, dict):
                             tc_args = {}
                         if tc_name:
-                            parts.append(
-                                types.Part.from_function_call(
-                                    name=tc_name, args=tc_args
-                                )
+                            fc_part = types.Part.from_function_call(
+                                name=tc_name, args=tc_args
                             )
+                            # Thinking-capable Gemini models require the original
+                            # thought_signature to be echoed back on functionCall
+                            # parts after a tool response, or the next request
+                            # fails with INVALID_ARGUMENT.
+                            sig = _decode_thought_signature(tc.get("thought_signature"))
+                            if sig:
+                                fc_part.thought_signature = sig
+                            parts.append(fc_part)
                     if not parts:
                         parts.append(types.Part.from_text(text=" "))
                     gemini_messages.append(types.Content(role="model", parts=parts))
@@ -310,6 +339,9 @@ class GeminiClient(GenAIClient):
                                 "id": part.function_call.name or "",
                                 "name": part.function_call.name or "",
                                 "arguments": arguments,
+                                "thought_signature": _encode_thought_signature(
+                                    getattr(part, "thought_signature", None)
+                                ),
                             }
                         )
 
@@ -369,11 +401,14 @@ class GeminiClient(GenAIClient):
         messages: list[dict[str, Any]],
         tools: Optional[list[dict[str, Any]]] = None,
         tool_choice: Optional[str] = "auto",
+        enable_thinking: Optional[bool] = None,
     ) -> AsyncGenerator[tuple[str, Any], None]:
         """
         Stream chat with tools; yields content deltas then final message.
 
         Implements streaming function calling/tool usage for Gemini models.
+        ``enable_thinking`` is accepted for interface parity; Gemini configures
+        thinking at the model level, so it is ignored here.
         """
         try:
             # Convert messages to Gemini format
@@ -415,11 +450,17 @@ class GeminiClient(GenAIClient):
                         if not isinstance(tc_args, dict):
                             tc_args = {}
                         if tc_name:
-                            parts.append(
-                                types.Part.from_function_call(
-                                    name=tc_name, args=tc_args
-                                )
+                            fc_part = types.Part.from_function_call(
+                                name=tc_name, args=tc_args
                             )
+                            # Thinking-capable Gemini models require the original
+                            # thought_signature to be echoed back on functionCall
+                            # parts after a tool response, or the next request
+                            # fails with INVALID_ARGUMENT.
+                            sig = _decode_thought_signature(tc.get("thought_signature"))
+                            if sig:
+                                fc_part.thought_signature = sig
+                            parts.append(fc_part)
                     if not parts:
                         parts.append(types.Part.from_text(text=" "))
                     gemini_messages.append(types.Content(role="model", parts=parts))
@@ -585,6 +626,7 @@ class GeminiClient(GenAIClient):
                                     "id": tool_call_id,
                                     "name": tool_call_name,
                                     "arguments": "",
+                                    "thought_signature": None,
                                 }
 
                             # Accumulate arguments
@@ -595,6 +637,13 @@ class GeminiClient(GenAIClient):
                                     else str(arguments)
                                 )
 
+                            # Capture latest thought_signature for this call
+                            chunk_sig = getattr(part, "thought_signature", None)
+                            if chunk_sig:
+                                tool_calls_by_index[found_index][
+                                    "thought_signature"
+                                ] = chunk_sig
+
             # Build final message
             full_content = "".join(content_parts).strip() or None
             full_reasoning = "".join(reasoning_parts).strip() or None
@@ -615,6 +664,9 @@ class GeminiClient(GenAIClient):
                             "id": tc["id"],
                             "name": tc["name"],
                             "arguments": parsed_args,
+                            "thought_signature": _encode_thought_signature(
+                                tc.get("thought_signature")
+                            ),
                         }
                     )
                 finish_reason = "tool_calls"

frigate/genai/plugins/openai.py

@@ -309,11 +309,15 @@ class OpenAIClient(GenAIClient):
         messages: list[dict[str, Any]],
         tools: Optional[list[dict[str, Any]]] = None,
         tool_choice: Optional[str] = "auto",
+        enable_thinking: Optional[bool] = None,
     ) -> AsyncGenerator[tuple[str, Any], None]:
         """
         Stream chat with tools; yields content deltas then final message.
 
         Implements streaming function calling/tool usage for OpenAI models.
+        The OpenAI chat completions API does not expose a per-request thinking
+        toggle, so ``enable_thinking`` is accepted for interface parity and
+        ignored.
         """
         try:
             openai_tool_choice = None

frigate/genai/utils.py

@@ -69,6 +69,14 @@ def build_assistant_message_for_conversation(
                     "name": tc["name"],
                     "arguments": json.dumps(tc.get("arguments") or {}),
                 },
+                # Gemini-only: opaque signature that must be echoed back on
+                # the same functionCall part in the next turn. Other providers
+                # do not set or read this.
+                **(
+                    {"thought_signature": tc["thought_signature"]}
+                    if tc.get("thought_signature")
+                    else {}
+                ),
             }
             for tc in tool_calls_raw
         ]

frigate/object_detection/base.py

@@ -167,8 +167,9 @@ class DetectorRunner(FrigateProcess):
 
             # detect and send the output
             self.start_time.value = datetime.datetime.now().timestamp()
+            mono_start = time.monotonic()
             detections = object_detector.detect_raw(input_frame)
-            duration = datetime.datetime.now().timestamp() - self.start_time.value
+            duration = time.monotonic() - mono_start
             frame_manager.close(connection_id)
 
             if connection_id not in self.outputs:

frigate/ptz/autotrack.py

@@ -1331,6 +1331,8 @@ class PtzAutoTracker:
         return self.tracked_object[camera]["region"]
 
     def autotrack_object(self, camera: str, obj: TrackedObject):
+        if camera not in self.config.cameras:
+            return
         camera_config = self.config.cameras[camera]
 
         if camera_config.onvif.autotracking.enabled:

frigate/test/test_camera_maintainer.py

@@ -0,0 +1,79 @@
+"""Tests for CameraMaintainer SHM cleanup on camera remove.
+
+Regression coverage for the case where a camera is removed and then a
+new camera is added with the same name. Without unlinking the per-frame
+YUV SHM slots, the maintainer's frame_manager.create call hits
+FileExistsError and falls back to reopening the existing segment at the
+*old* size, which the new ffmpeg process then writes mismatched-size
+frames into.
+"""
+
+import unittest
+from unittest.mock import MagicMock, patch
+
+from frigate.camera.maintainer import CameraMaintainer
+
+
+class TestMaintainerUnlinkFrameSlotsOnRemove(unittest.TestCase):
+    def _make_maintainer(self) -> CameraMaintainer:
+        """Build a maintainer without invoking __init__ (avoids needing real
+        FrigateConfig, queues, multiprocessing manager, etc.). We're only
+        exercising the SHM-cleanup helper, so the surrounding init is
+        irrelevant."""
+        maintainer = CameraMaintainer.__new__(CameraMaintainer)
+        maintainer.frame_manager = MagicMock()
+        return maintainer
+
+    def test_unlinks_only_segments_with_matching_prefix(self) -> None:
+        maintainer = self._make_maintainer()
+        maintainer.frame_manager.shm_store = {
+            "front_frame0": object(),
+            "front_frame1": object(),
+            "front_frame2": object(),
+            # Different camera; must not be touched.
+            "side_frame0": object(),
+            # Detector input/output buffers are sized by the model and
+            # cached by the long-lived DetectorRunner — must not be
+            # touched even when their owning camera is removed.
+            "front": object(),
+            "out-front": object(),
+        }
+
+        # __name-mangled access from outside the class.
+        maintainer._CameraMaintainer__unlink_camera_frame_slots("front")
+
+        deleted = [c.args[0] for c in maintainer.frame_manager.delete.call_args_list]
+        self.assertEqual(
+            sorted(deleted),
+            ["front_frame0", "front_frame1", "front_frame2"],
+        )
+
+    def test_handles_camera_with_no_slots(self) -> None:
+        """Cameras that were removed before any frame slot was ever
+        created (e.g. cancelled during preparing_clip) should be a no-op."""
+        maintainer = self._make_maintainer()
+        maintainer.frame_manager.shm_store = {"other_frame0": object()}
+
+        maintainer._CameraMaintainer__unlink_camera_frame_slots("front")
+
+        maintainer.frame_manager.delete.assert_not_called()
+
+    def test_swallows_delete_errors(self) -> None:
+        """Unlink failures shouldn't abort the remove loop — best-effort."""
+        maintainer = self._make_maintainer()
+        maintainer.frame_manager.shm_store = {
+            "front_frame0": object(),
+            "front_frame1": object(),
+        }
+        maintainer.frame_manager.delete.side_effect = OSError("simulated")
+
+        # Both slots are attempted; the OSError on the first doesn't
+        # prevent the second from being tried.
+        with patch("frigate.camera.maintainer.logger"):
+            maintainer._CameraMaintainer__unlink_camera_frame_slots("front")
+
+        self.assertEqual(maintainer.frame_manager.delete.call_count, 2)
+
+
+if __name__ == "__main__":
+    unittest.main()

frigate/test/test_shared_memory_frame_manager.py

@@ -0,0 +1,156 @@
+"""Tests for SharedMemoryFrameManager cache invalidation.
+
+Covers the case where a SHM segment is unlinked and recreated at a
+different size across a camera add/remove cycle while a long-lived
+in-process cache (e.g. TrackedObjectProcessor) still holds a ref to
+the old, smaller segment.
+"""
+
+import unittest
+from types import SimpleNamespace
+from unittest.mock import patch
+
+import numpy as np
+
+from frigate.util.image import SharedMemoryFrameManager
+
+
+def _fake_shm(size: int) -> SimpleNamespace:
+    """A minimal stand-in for UntrackedSharedMemory with .size and .buf."""
+    return SimpleNamespace(size=size, buf=bytearray(size), close=lambda: None)
+
+
+class TestSharedMemoryFrameManagerGet(unittest.TestCase):
+    def test_get_reopens_when_cached_segment_is_smaller_than_shape(self) -> None:
+        """A cached ref to an older smaller segment must be dropped and the
+        current (correctly sized) segment reopened. Without this, np.ndarray
+        would raise "buffer is too small for requested array" when the
+        in-memory cache pointed at an old SHM after a same-name resize."""
+        manager = SharedMemoryFrameManager()
+
+        small = _fake_shm(size=100)
+        current = _fake_shm(size=2_500)
+        manager.shm_store["cam_frame0"] = small
+
+        with patch("frigate.util.image.UntrackedSharedMemory", return_value=current):
+            arr = manager.get("cam_frame0", (50, 50))
+
+        self.assertIsNotNone(arr)
+        self.assertEqual(arr.shape, (50, 50))
+        self.assertIs(manager.shm_store["cam_frame0"], current)
+
+    def test_get_reopens_when_cached_segment_is_larger_than_shape(self) -> None:
+        """Symmetric to the smaller-cache case: when detect resolution drops,
+        the SHM is unlinked and recreated at a smaller size. A cached ref to
+        the old, larger segment still satisfies any size check but points at
+        an orphaned inode whose stale bytes get reinterpreted at the new
+        shape — producing miscolored, distorted YUV frames downstream. Drop
+        the cache so we reopen by name and bind to the current segment."""
+        manager = SharedMemoryFrameManager()
+
+        old_large = _fake_shm(size=10_000)
+        current = _fake_shm(size=2_500)
+        manager.shm_store["cam_frame0"] = old_large
+
+        with patch("frigate.util.image.UntrackedSharedMemory", return_value=current):
+            arr = manager.get("cam_frame0", (50, 50))
+
+        self.assertIsNotNone(arr)
+        self.assertEqual(arr.shape, (50, 50))
+        self.assertIs(manager.shm_store["cam_frame0"], current)
+
+    def test_get_keeps_cached_segment_when_size_matches(self) -> None:
+        """Don't pay the reopen cost when the cached ref is the right size."""
+        manager = SharedMemoryFrameManager()
+
+        cached = _fake_shm(size=2_500)
+        manager.shm_store["cam_frame0"] = cached
+
+        with patch("frigate.util.image.UntrackedSharedMemory") as untracked_shm_cls:
+            arr = manager.get("cam_frame0", (50, 50))
+            untracked_shm_cls.assert_not_called()
+
+        self.assertIsNotNone(arr)
+        self.assertIs(manager.shm_store["cam_frame0"], cached)
+
+    def test_get_opens_fresh_when_no_cache_entry(self) -> None:
+        manager = SharedMemoryFrameManager()
+        fresh = _fake_shm(size=2_500)
+
+        with patch("frigate.util.image.UntrackedSharedMemory", return_value=fresh):
+            arr = manager.get("cam_frame0", (50, 50))
+
+        self.assertIsNotNone(arr)
+        self.assertIs(manager.shm_store["cam_frame0"], fresh)
+
+    def test_get_returns_none_when_segment_missing(self) -> None:
+        manager = SharedMemoryFrameManager()
+
+        with patch(
+            "frigate.util.image.UntrackedSharedMemory",
+            side_effect=FileNotFoundError,
+        ):
+            arr = manager.get("cam_frame0", (50, 50))
+
+        self.assertIsNone(arr)
+
+    def test_get_returns_none_when_reopened_segment_is_still_too_small(self) -> None:
+        """Race during a same-name SHM recreate: cache is stale, we reopen
+        by name, but the maintainer hasn't allocated the new segment yet —
+        the reopened ref is also too small. Skip the frame (return None)
+        rather than crash on np.ndarray."""
+        manager = SharedMemoryFrameManager()
+
+        small_cached = _fake_shm(size=100)
+        still_small_after_reopen = _fake_shm(size=100)
+        manager.shm_store["cam_frame0"] = small_cached
+
+        with patch(
+            "frigate.util.image.UntrackedSharedMemory",
+            return_value=still_small_after_reopen,
+        ):
+            arr = manager.get("cam_frame0", (50, 50))
+
+        self.assertIsNone(arr)
+        # Don't cache the too-small reopened ref — next call will re-open
+        # once the maintainer has finished recreating the segment.
+        self.assertNotIn("cam_frame0", manager.shm_store)
+
+    def test_get_handles_n_dimensional_shape(self) -> None:
+        """np.prod must be used (not raw multiplication) for tuple shapes."""
+        manager = SharedMemoryFrameManager()
+        # YUV-shaped frame: (height * 3/2, width) for 1920x1080 = 3,110,400
+        big_enough = _fake_shm(size=3_110_400)
+        manager.shm_store["cam_frame0"] = big_enough
+
+        with patch("frigate.util.image.UntrackedSharedMemory") as untracked_shm_cls:
+            arr = manager.get("cam_frame0", (1620, 1920))
+            untracked_shm_cls.assert_not_called()
+
+        self.assertIsNotNone(arr)
+        self.assertEqual(arr.shape, (1620, 1920))
+
+
+class TestSharedMemoryFrameManagerGetRecreatesLargerSegment(unittest.TestCase):
+    """End-to-end-style: simulates the full unlink-and-recreate cycle."""
+
+    def test_segment_grows_then_get_succeeds(self) -> None:
+        manager = SharedMemoryFrameManager()
+
+        # Phase 1: existing camera at 320x240 YUV — 320 * 240 * 1.5 = 115_200
+        small = _fake_shm(size=115_200)
+        manager.shm_store["cam_frame0"] = small
+        arr_small = np.ndarray((360, 320), dtype=np.uint8, buffer=small.buf)
+        self.assertEqual(arr_small.shape, (360, 320))
+
+        # Phase 2: restart at 1920x1080 — new SHM segment, larger size.
+        large = _fake_shm(size=3_110_400)
+        with patch("frigate.util.image.UntrackedSharedMemory", return_value=large):
+            arr_large = manager.get("cam_frame0", (1620, 1920))
+
+        self.assertIsNotNone(arr_large)
+        self.assertEqual(arr_large.shape, (1620, 1920))
+
+
+if __name__ == "__main__":
+    unittest.main()

frigate/util/config.py

@@ -788,6 +788,45 @@ def apply_section_update(camera_config, section: str, update: dict) -> Optional[
                 )
             camera_config.objects = new_objects
 
+        elif section == "detect":
+            # apply detect first so frame_shape reflects the new resolution
+            # before we rebuild mask-dependent runtime configs below
+            merged = deep_merge(current.model_dump(), update, override=True)
+            camera_config.detect = current.__class__.model_validate(merged)
+
+            new_frame_shape = camera_config.frame_shape
+
+            # rebuild motion's rasterized_mask at the new frame_shape
+            if camera_config.motion is not None:
+                camera_config.motion = RuntimeMotionConfig(
+                    frame_shape=new_frame_shape,
+                    **camera_config.motion.model_dump(exclude_unset=True),
+                )
+
+            # rebuild per-object filter masks at the new frame_shape
+            for obj_name, filt in camera_config.objects.filters.items():
+                merged_mask = dict(filt.mask)
+                if camera_config.objects.mask:
+                    for gid, gmask in camera_config.objects.mask.items():
+                        merged_mask[f"global_{gid}"] = gmask
+
+                camera_config.objects.filters[obj_name] = RuntimeFilterConfig(
+                    frame_shape=new_frame_shape,
+                    mask=merged_mask,
+                    **filt.model_dump(exclude_unset=True, exclude={"mask", "raw_mask"}),
+                )
+
+            # Regenerate zone contours and per-zone filter masks at the new
+            # frame_shape so zone outlines and membership stay relative
+            for zone in camera_config.zones.values():
+                if zone.filters:
+                    for zone_obj_name, zone_filter in zone.filters.items():
+                        zone.filters[zone_obj_name] = RuntimeFilterConfig(
+                            frame_shape=new_frame_shape,
+                            **zone_filter.model_dump(exclude_unset=True),
+                        )
+                zone.generate_contour(new_frame_shape)
+
         else:
             merged = deep_merge(current.model_dump(), update, override=True)
             setattr(camera_config, section, current.__class__.model_validate(merged))

frigate/util/image.py

@@ -1089,10 +1089,25 @@ class SharedMemoryFrameManager(FrameManager):
 
     def get(self, name: str, shape) -> Optional[np.ndarray]:
         try:
-            if name in self.shm_store:
-                shm = self.shm_store[name]
-            else:
+            required = int(np.prod(shape))
+            shm = self.shm_store.get(name)
+            if shm is not None and shm.size != required:
+                # stale cached ref from a same-name recreate — drop and reopen
+                try:
+                    shm.close()
+                except Exception:
+                    pass
+                self.shm_store.pop(name, None)
+                shm = None
+            if shm is None:
                 shm = UntrackedSharedMemory(name=name)
+                if shm.size != required:
+                    # mid-recreate: OS segment doesn't match shape yet; skip
+                    try:
+                        shm.close()
+                    except Exception:
+                        pass
+                    return None
                 self.shm_store[name] = shm
             return np.ndarray(shape, dtype=np.uint8, buffer=shm.buf)
         except FileNotFoundError:

web/public/locales/en/config/validation.json

@@ -28,5 +28,8 @@
       "detectRequired": "At least one input stream must be assigned the 'detect' role.",
       "hwaccelDetectOnly": "Only the input stream with the detect role can define hardware acceleration arguments."
     }
+  },
+  "detect": {
+    "dimensionMustBeEven": "Must be an even number."
   }
 }

web/public/locales/en/views/settings.json

@@ -484,11 +484,15 @@
       "reorderHandle": "Drag to reorder",
       "saving": "Saving…",
       "saved": "Saved",
-      "friendlyName": {
-        "edit": "Edit camera display name",
-        "title": "Edit Display Name",
-        "description": "Set the friendly name shown for this camera throughout the Frigate UI. Leave blank to use the camera ID.",
-        "rename": "Rename"
+      "details": {
+        "edit": "Edit camera details",
+        "title": "Edit Camera Details",
+        "description": "Update the display name and external URL used for this camera throughout the Frigate UI.",
+        "friendlyNameLabel": "Display Name",
+        "friendlyNameHelp": "Friendly name shown for this camera throughout the Frigate UI. Leave blank to use the camera ID.",
+        "webuiUrlLabel": "Camera Web UI URL",
+        "webuiUrlHelp": "URL to visit the camera's web UI directly from the Debug view. Leave blank to disable the link.",
+        "webuiUrlInvalid": "Must be a valid URL (e.g., https://example.com)."
       }
     },
     "cameraConfig": {
@@ -1539,6 +1543,9 @@
       "builtIn": "Built-in Models",
       "genaiProviders": "GenAI Providers"
     },
+    "semanticSearchModelSize": {
+      "notApplicable": "Not applicable for GenAI providers"
+    },
     "review": {
       "title": "Review Settings"
     },
@@ -1787,7 +1794,9 @@
     },
     "detect": {
       "fpsGreaterThanFive": "Setting the detect FPS higher than 5 is not recommended. Higher values may cause performance issues and will not provide any benefit.",
-      "disabled": "Object detection is disabled. Snapshots, review items, and enrichments such as face recognition, license plate recognition, and Generative AI will not function."
+      "disabled": "Object detection is disabled. Snapshots, review items, and enrichments such as face recognition, license plate recognition, and Generative AI will not function.",
+      "resolutionShouldBeMultipleOfFour": "For best results, detect width and height should be multiples of 4. Other even values may produce visual artifacts or slight distortion in the detect stream.",
+      "aspectRatioMismatch": "The width and height you've entered don't match the aspect ratio of your current detect resolution. This may produce a stretched or distorted image."
     },
     "objects": {
       "genaiNoDescriptionsProvider": "You must configure a GenAI provider with the 'descriptions' role for descriptions to be generated."

web/src/components/config-form/FieldMessagesContext.ts

@@ -0,0 +1,13 @@
+import { createContext } from "react";
+import type { FieldConditionalMessage } from "./section-configs/types";
+
+// Provides currently-active field messages to FieldTemplate without going
+// through RJSF's per-field uiSchema. RJSF caches state.uiSchema across renders
+// in a way that can leave stale ui:messages attached to a field when the
+// triggering condition flips back to false (see processPendingChange in
+// @rjsf/core Form.js — formData is updated immediately, uiSchema is not).
+// useContext re-runs consumers directly on provider value change, sidestepping
+// that staleness.
+export const FieldMessagesContext = createContext<FieldConditionalMessage[]>(
+  [],
+);

web/src/components/config-form/LiveFormDataContext.ts

@@ -0,0 +1,13 @@
+import { createContext } from "react";
+import type { ConfigSectionData } from "@/types/configForm";
+
+// Mirrors the current section's in-flight form data so widgets can react
+// to changes that RJSF wouldn't otherwise re-render them for. RJSF's
+// Form memoizes SchemaField via deep equality and, in some transitions
+// (notably reverting a field to its saved value), can skip re-rendering
+// a widget even though the form data it depends on changed. useContext
+// re-runs consumers directly on every provider value update, sidestepping
+// that.
+export const LiveFormDataContext = createContext<ConfigSectionData | null>(
+  null,
+);

web/src/components/config-form/section-configs/detect.ts

@@ -11,6 +11,50 @@ const detect: SectionConfigOverrides = {
         condition: (ctx) =>
           ctx.level === "camera" && ctx.formData?.enabled === false,
       },
+      {
+        key: "detect-resolution-not-multiple-of-four",
+        messageKey: "configMessages.detect.resolutionShouldBeMultipleOfFour",
+        severity: "warning",
+        condition: (ctx) => {
+          const width = ctx.formData?.width as number | null | undefined;
+          const height = ctx.formData?.height as number | null | undefined;
+          const isEvenButNotFour = (v: unknown) =>
+            typeof v === "number" && v % 2 === 0 && v % 4 !== 0;
+          return isEvenButNotFour(width) || isEvenButNotFour(height);
+        },
+      },
+      {
+        key: "detect-aspect-ratio-mismatch",
+        messageKey: "configMessages.detect.aspectRatioMismatch",
+        severity: "warning",
+        condition: (ctx) => {
+          const newWidth = ctx.formData?.width as number | null | undefined;
+          const newHeight = ctx.formData?.height as number | null | undefined;
+          if (typeof newWidth !== "number" || typeof newHeight !== "number") {
+            return false;
+          }
+          const saved =
+            ctx.level === "camera"
+              ? ctx.fullCameraConfig?.detect
+              : ctx.fullConfig?.detect;
+          const savedWidth = saved?.width;
+          const savedHeight = saved?.height;
+          if (
+            typeof savedWidth !== "number" ||
+            typeof savedHeight !== "number" ||
+            savedWidth <= 0 ||
+            savedHeight <= 0
+          ) {
+            return false;
+          }
+          if (newWidth === savedWidth && newHeight === savedHeight) {
+            return false;
+          }
+          const newRatio = newWidth / newHeight;
+          const savedRatio = savedWidth / savedHeight;
+          return Math.abs(newRatio - savedRatio) > 0.01;
+        },
+      },
     ],
     fieldMessages: [
       {
@@ -72,6 +116,25 @@ const detect: SectionConfigOverrides = {
       "max_disappeared",
     ],
   },
+  replay: {
+    restartRequired: [],
+    fieldOrder: ["width", "height", "fps"],
+    fieldGroups: {
+      resolution: ["width", "height", "fps"],
+    },
+    hiddenFields: [
+      "enabled",
+      "enabled_in_config",
+      "min_initialized",
+      "max_disappeared",
+      "annotation_offset",
+      "stationary",
+      "interval",
+      "threshold",
+      "max_frames",
+    ],
+    advancedFields: [],
+  },
 };
 
 export default detect;

web/src/components/config-form/section-configs/semantic_search.ts

@@ -35,6 +35,7 @@ const semanticSearch: SectionConfigOverrides = {
         "ui:widget": "semanticSearchModel",
       },
       model_size: {
+        "ui:widget": "semanticSearchModelSize",
         "ui:options": { size: "xs", enumI18nPrefix: "modelSize" },
       },
     },

web/src/components/config-form/section-validations/detect.ts

@@ -0,0 +1,36 @@
+import type { FormValidation } from "@rjsf/utils";
+import type { TFunction } from "i18next";
+import { isJsonObject } from "@/lib/utils";
+import type { JsonObject } from "@/types/configForm";
+
+export function validateDetectDimensions(
+  formData: unknown,
+  errors: FormValidation,
+  t: TFunction,
+): FormValidation {
+  if (!isJsonObject(formData as JsonObject)) {
+    return errors;
+  }
+
+  const data = formData as JsonObject;
+  const width = data.width;
+  const height = data.height;
+
+  const widthErrors = errors.width as
+    | { addError?: (message: string) => void }
+    | undefined;
+  const heightErrors = errors.height as
+    | { addError?: (message: string) => void }
+    | undefined;
+
+  const message = t("detect.dimensionMustBeEven", { ns: "config/validation" });
+
+  if (typeof width === "number" && width % 2 !== 0) {
+    widthErrors?.addError?.(message);
+  }
+  if (typeof height === "number" && height % 2 !== 0) {
+    heightErrors?.addError?.(message);
+  }
+
+  return errors;
+}

web/src/components/config-form/section-validations/index.ts

@@ -1,5 +1,6 @@
 import type { FormValidation } from "@rjsf/utils";
 import type { TFunction } from "i18next";
+import { validateDetectDimensions } from "./detect";
 import { validateFfmpegInputRoles } from "./ffmpeg";
 import { validateProxyRoleHeader } from "./proxy";
 
@@ -19,6 +20,10 @@ export function getSectionValidation({
   level,
   t,
 }: SectionValidationOptions): SectionValidation | undefined {
+  if (sectionPath === "detect") {
+    return (formData, errors) => validateDetectDimensions(formData, errors, t);
+  }
+
   if (sectionPath === "ffmpeg" && level === "camera") {
     return (formData, errors) => validateFfmpegInputRoles(formData, errors, t);
   }

web/src/components/config-form/sections/BaseSection.tsx

@@ -86,6 +86,8 @@ import type {
 } from "../section-configs/types";
 import { useConfigMessages } from "@/hooks/use-config-messages";
 import { ConfigMessageBanner } from "../ConfigMessageBanner";
+import { FieldMessagesContext } from "../FieldMessagesContext";
+import { LiveFormDataContext } from "../LiveFormDataContext";
 
 export interface SectionConfig {
   /** Field ordering within the section */
@@ -627,44 +629,6 @@ export function ConfigSection({
     messageContext,
   );
 
-  // Merge field-level conditional messages into uiSchema
-  const effectiveUiSchema = useMemo(() => {
-    if (activeFieldMessages.length === 0) return sectionConfig.uiSchema;
-    const merged = { ...(sectionConfig.uiSchema ?? {}) };
-    for (const msg of activeFieldMessages) {
-      const segments = msg.field.split(".");
-      // Navigate to the nested uiSchema node, shallow-cloning along the way
-      let node = merged;
-      for (let i = 0; i < segments.length - 1; i++) {
-        const seg = segments[i];
-        node[seg] = { ...(node[seg] as Record<string, unknown>) };
-        node = node[seg] as Record<string, unknown>;
-      }
-      const leafKey = segments[segments.length - 1];
-      const existing = node[leafKey] as Record<string, unknown> | undefined;
-      const existingMessages = ((existing?.["ui:messages"] as unknown[]) ??
-        []) as Array<{
-        key: string;
-        messageKey: string;
-        severity: string;
-        position?: string;
-      }>;
-      node[leafKey] = {
-        ...existing,
-        "ui:messages": [
-          ...existingMessages,
-          {
-            key: msg.key,
-            messageKey: msg.messageKey,
-            severity: msg.severity,
-            position: msg.position ?? "before",
-          },
-        ],
-      };
-    }
-    return merged;
-  }, [sectionConfig.uiSchema, activeFieldMessages]);
-
   const currentOverrides = useMemo(() => {
     if (!currentFormData || typeof currentFormData !== "object") {
       return undefined;
@@ -1034,59 +998,65 @@ export function ConfigSection({
   const sectionContent = (
     <div className="space-y-6">
       <ConfigMessageBanner messages={activeMessages} />
-      <ConfigForm
-        key={formKey}
-        schema={modifiedSchema}
-        formData={currentFormData}
-        onChange={handleChange}
-        onValidationChange={setHasValidationErrors}
-        fieldOrder={sectionConfig.fieldOrder}
-        fieldGroups={sectionConfig.fieldGroups}
-        hiddenFields={effectiveHiddenFields}
-        advancedFields={sectionConfig.advancedFields}
-        liveValidate={sectionConfig.liveValidate}
-        uiSchema={effectiveUiSchema}
-        disabled={disabled || isSaving}
-        readonly={readonly}
-        showSubmit={false}
-        i18nNamespace={configNamespace}
-        customValidate={customValidate}
-        formContext={{
-          level: effectiveLevel,
-          cameraName,
-          globalValue,
-          cameraValue,
-          hasChanges,
-          extraHasChanges,
-          setExtraHasChanges,
-          overrides: uiOverrides as JsonValue | undefined,
-          formData: currentFormData as ConfigSectionData,
-          baselineFormData: effectiveBaselineFormData as ConfigSectionData,
-          pendingDataBySection,
-          onPendingDataChange,
-          onFormDataChange: (data: ConfigSectionData) => handleChange(data),
-          // For widgets that need access to full camera config (e.g., zone names)
-          fullCameraConfig:
-            effectiveLevel === "camera" && cameraName
-              ? config?.cameras?.[cameraName]
-              : undefined,
-          fullConfig: config,
-          // When rendering camera-level sections, provide the section path so
-          // field templates can look up keys under the `config/cameras` namespace
-          // When using a consolidated global namespace, keys are nested
-          // under the section name (e.g., `audio.label`) so provide the
-          // section prefix to templates so they can attempt `${section}.${field}` lookups.
-          sectionI18nPrefix: sectionPath,
-          t,
-          renderers: wrappedRenderers,
-          sectionDocs: sectionConfig.sectionDocs,
-          fieldDocs: sectionConfig.fieldDocs,
-          hiddenFields: effectiveHiddenFields,
-          restartRequired: sectionConfig.restartRequired,
-          requiresRestart,
-          isProfile: !!profileName,
-        }}
-      />
+      <FieldMessagesContext.Provider value={activeFieldMessages}>
+        <LiveFormDataContext.Provider
+          value={(currentFormData as ConfigSectionData | null) ?? null}
+        >
+          <ConfigForm
+            key={formKey}
+            schema={modifiedSchema}
+            formData={currentFormData}
+            onChange={handleChange}
+            onValidationChange={setHasValidationErrors}
+            fieldOrder={sectionConfig.fieldOrder}
+            fieldGroups={sectionConfig.fieldGroups}
+            hiddenFields={effectiveHiddenFields}
+            advancedFields={sectionConfig.advancedFields}
+            liveValidate={sectionConfig.liveValidate}
+            uiSchema={sectionConfig.uiSchema}
+            disabled={disabled || isSaving}
+            readonly={readonly}
+            showSubmit={false}
+            i18nNamespace={configNamespace}
+            customValidate={customValidate}
+            formContext={{
+              level: effectiveLevel,
+              cameraName,
+              globalValue,
+              cameraValue,
+              hasChanges,
+              extraHasChanges,
+              setExtraHasChanges,
+              overrides: uiOverrides as JsonValue | undefined,
+              formData: currentFormData as ConfigSectionData,
+              baselineFormData: effectiveBaselineFormData as ConfigSectionData,
+              pendingDataBySection,
+              onPendingDataChange,
+              onFormDataChange: (data: ConfigSectionData) => handleChange(data),
+              // For widgets that need access to full camera config (e.g., zone names)
+              fullCameraConfig:
+                effectiveLevel === "camera" && cameraName
+                  ? config?.cameras?.[cameraName]
+                  : undefined,
+              fullConfig: config,
+              // When rendering camera-level sections, provide the section path so
+              // field templates can look up keys under the `config/cameras` namespace
+              // When using a consolidated global namespace, keys are nested
+              // under the section name (e.g., `audio.label`) so provide the
+              // section prefix to templates so they can attempt `${section}.${field}` lookups.
+              sectionI18nPrefix: sectionPath,
+              t,
+              renderers: wrappedRenderers,
+              sectionDocs: sectionConfig.sectionDocs,
+              fieldDocs: sectionConfig.fieldDocs,
+              hiddenFields: effectiveHiddenFields,
+              restartRequired: sectionConfig.restartRequired,
+              requiresRestart,
+              isProfile: !!profileName,
+            }}
+          />
+        </LiveFormDataContext.Provider>
+      </FieldMessagesContext.Provider>
 
       {!embedded && (
         <div
@@ -1288,7 +1258,12 @@ export function ConfigSection({
             <CollapsibleTrigger asChild>
               <div className="flex cursor-pointer items-center justify-between">
                 <div className="flex items-center gap-3">
-                  <Heading as="h4">{title}</Heading>
+                  <Heading
+                    as="h4"
+                    className={level === "replay" ? "text-base" : undefined}
+                  >
+                    {title}
+                  </Heading>
                   {showOverrideIndicator &&
                     effectiveLevel === "camera" &&
                     (profileOverridesSection || isOverridden) &&

web/src/components/config-form/theme/frigateTheme.ts

@@ -31,6 +31,7 @@ import { TimezoneSelectWidget } from "./widgets/TimezoneSelectWidget";
 import { CameraPathWidget } from "./widgets/CameraPathWidget";
 import { OptionalFieldWidget } from "./widgets/OptionalFieldWidget";
 import { SemanticSearchModelWidget } from "./widgets/SemanticSearchModelWidget";
+import { SemanticSearchModelSizeWidget } from "./widgets/SemanticSearchModelSizeWidget";
 import { OnvifProfileWidget } from "./widgets/OnvifProfileWidget";
 
 import { FieldTemplate } from "./templates/FieldTemplate";
@@ -86,6 +87,7 @@ export const frigateTheme: FrigateTheme = {
     timezoneSelect: TimezoneSelectWidget,
     optionalField: OptionalFieldWidget,
     semanticSearchModel: SemanticSearchModelWidget,
+    semanticSearchModelSize: SemanticSearchModelSizeWidget,
     onvifProfile: OnvifProfileWidget,
   },
   templates: {

web/src/components/config-form/theme/templates/FieldTemplate.tsx

@@ -5,8 +5,9 @@ import {
   getUiOptions,
   ADDITIONAL_PROPERTY_FLAG,
 } from "@rjsf/utils";
-import { ComponentType, ReactNode } from "react";
+import { ComponentType, ReactNode, useContext } from "react";
 import { isValidElement } from "react";
+import { FieldMessagesContext } from "../../FieldMessagesContext";
 import { Label } from "@/components/ui/label";
 import { cn } from "@/lib/utils";
 import { useTranslation } from "react-i18next";
@@ -95,6 +96,7 @@ export function FieldTemplate(props: FieldTemplateProps) {
     "views/settings",
   ]);
   const { getLocaleDocUrl } = useDocDomain();
+  const allFieldMessages = useContext(FieldMessagesContext);
 
   if (hidden) {
     return <div className="hidden">{children}</div>;
@@ -384,21 +386,15 @@ export function FieldTemplate(props: FieldTemplateProps) {
   const beforeContent = renderCustom(beforeSpec);
   const afterContent = renderCustom(afterSpec);
 
-  // Render conditional field messages from ui:messages
-  const fieldMessageSpecs = uiSchema?.["ui:messages"] as
-    | Array<{
-        key: string;
-        messageKey: string;
-        severity: string;
-        position?: string;
-      }>
-    | undefined;
-  const beforeMessages = fieldMessageSpecs?.filter(
+  // Read field-level conditional messages from FieldMessagesContext
+  const fieldPathStr = pathSegments.join(".");
+  const fieldMessageSpecs = allFieldMessages.filter(
+    (m) => m.field === fieldPathStr,
+  );
+  const beforeMessages = fieldMessageSpecs.filter(
     (m) => (m.position ?? "before") === "before",
   );
-  const afterMessages = fieldMessageSpecs?.filter(
-    (m) => m.position === "after",
-  );
+  const afterMessages = fieldMessageSpecs.filter((m) => m.position === "after");
   const beforeMessagesContent =
     beforeMessages && beforeMessages.length > 0 ? (
       <div className="space-y-2">

web/src/components/config-form/theme/widgets/SemanticSearchModelSizeWidget.tsx

@@ -0,0 +1,57 @@
+// Disables model_size and shows "N/A" when a GenAI provider is selected.
+// Reads model via LiveFormDataContext so it re-runs even when RJSF's
+// SchemaField memoization would skip this widget.
+import type { WidgetProps } from "@rjsf/utils";
+import { useContext, useEffect } from "react";
+import { useTranslation } from "react-i18next";
+import {
+  Select,
+  SelectContent,
+  SelectTrigger,
+  SelectValue,
+} from "@/components/ui/select";
+import { LiveFormDataContext } from "../../LiveFormDataContext";
+import { getSizedFieldClassName } from "../utils";
+import { SelectWidget } from "./SelectWidget";
+
+export function SemanticSearchModelSizeWidget(props: WidgetProps) {
+  const { t } = useTranslation(["views/settings"]);
+  const liveFormData = useContext(LiveFormDataContext);
+  const model = liveFormData?.model;
+  const isProvider =
+    typeof model === "string" &&
+    model !== "" &&
+    model !== "jinav1" &&
+    model !== "jinav2";
+
+  // Clear model_size while on a provider (buildOverrides converts to ""
+  // which the backend treats as "remove"). Restore the schema default
+  // when returning to a Jina model so the field isn't left empty.
+  const { value, onChange, schema } = props;
+  const schemaDefault = schema?.default as string | undefined;
+  useEffect(() => {
+    if (isProvider && value !== undefined) {
+      onChange(undefined);
+    } else if (!isProvider && value === undefined && schemaDefault) {
+      onChange(schemaDefault);
+    }
+  }, [isProvider, value, onChange, schemaDefault]);
+
+  if (isProvider) {
+    const fieldClassName = getSizedFieldClassName(props.options ?? {}, "sm");
+    return (
+      <Select value="" disabled>
+        <SelectTrigger className={fieldClassName}>
+          <SelectValue
+            placeholder={t("configForm.semanticSearchModelSize.notApplicable", {
+              defaultValue: "Not applicable for GenAI providers",
+            })}
+          />
+        </SelectTrigger>
+        <SelectContent />
+      </Select>
+    );
+  }
+
+  return <SelectWidget {...props} />;
+}

web/src/components/menu/SearchResultActions.tsx

@@ -1,5 +1,6 @@
 import { useState, ReactNode, useCallback } from "react";
 import { SearchResult } from "@/types/search";
+import { REVIEW_PADDING } from "@/types/review";
 import { FrigateConfig } from "@/types/frigateConfig";
 import { baseUrl } from "@/api/baseUrl";
 import { toast } from "sonner";
@@ -94,8 +95,8 @@ export default function SearchResultActions({
       axios
         .post("debug_replay/start", {
           camera: event.camera,
-          start_time: event.start_time,
-          end_time: event.end_time,
+          start_time: (event.start_time ?? 0) - REVIEW_PADDING,
+          end_time: (event.end_time ?? Date.now() / 1000) + REVIEW_PADDING,
         })
         .then((response) => {
           if (response.status === 202 || response.status === 200) {

web/src/components/overlay/DebugReplayConfigSheet.tsx

@@ -0,0 +1,120 @@
+import { useState } from "react";
+import { useTranslation } from "react-i18next";
+import { LuSettings } from "react-icons/lu";
+import useSWR from "swr";
+import ActivityIndicator from "@/components/indicators/activity-indicator";
+import { ConfigSectionTemplate } from "@/components/config-form/sections/ConfigSectionTemplate";
+import { Button } from "@/components/ui/button";
+import { PlatformAwareSheet } from "@/components/overlay/dialog/PlatformAwareDialog";
+import { useConfigSchema } from "@/hooks/use-config-schema";
+import type { FrigateConfig } from "@/types/frigateConfig";
+
+type DebugReplayConfigSheetProps = {
+  replayCamera: string | undefined;
+};
+
+export function DebugReplayConfigSheet({
+  replayCamera,
+}: DebugReplayConfigSheetProps) {
+  const { t } = useTranslation(["views/replay"]);
+  const configSchema = useConfigSchema();
+  const { data: config } = useSWR<FrigateConfig>("config", {
+    revalidateOnFocus: false,
+  });
+  const [open, setOpen] = useState(false);
+
+  return (
+    <PlatformAwareSheet
+      trigger={
+        <Button variant="outline" size="sm" className="flex items-center gap-2">
+          <LuSettings className="size-4" />
+          <span className="hidden md:inline">{t("page.configuration")}</span>
+        </Button>
+      }
+      title={t("page.configuration")}
+      titleClassName="text-lg font-semibold"
+      contentClassName="scrollbar-container flex flex-col gap-0 overflow-y-auto px-6 pb-6 sm:max-w-xl md:max-w-2xl xl:max-w-3xl"
+      content={
+        <>
+          <p className="mb-5 text-sm text-muted-foreground">
+            {t("page.configurationDesc")}
+          </p>
+          {configSchema == null ? (
+            <div className="flex h-40 items-center justify-center">
+              <ActivityIndicator />
+            </div>
+          ) : (
+            <div className="space-y-6">
+              <ConfigSectionTemplate
+                sectionKey="detect"
+                level="replay"
+                cameraName={replayCamera}
+                skipSave
+                noStickyButtons
+                requiresRestart={false}
+                collapsible
+                defaultCollapsed={false}
+                showTitle
+                showOverrideIndicator={false}
+              />
+              <ConfigSectionTemplate
+                sectionKey="motion"
+                level="replay"
+                cameraName={replayCamera}
+                skipSave
+                noStickyButtons
+                requiresRestart={false}
+                collapsible
+                defaultCollapsed={false}
+                showTitle
+                showOverrideIndicator={false}
+              />
+              <ConfigSectionTemplate
+                sectionKey="objects"
+                level="replay"
+                cameraName={replayCamera}
+                skipSave
+                noStickyButtons
+                requiresRestart={false}
+                collapsible
+                defaultCollapsed={false}
+                showTitle
+                showOverrideIndicator={false}
+              />
+              {config?.face_recognition?.enabled && (
+                <ConfigSectionTemplate
+                  sectionKey="face_recognition"
+                  level="replay"
+                  cameraName={replayCamera}
+                  skipSave
+                  noStickyButtons
+                  requiresRestart={false}
+                  collapsible
+                  defaultCollapsed={false}
+                  showTitle
+                  showOverrideIndicator={false}
+                />
+              )}
+              {config?.lpr?.enabled && (
+                <ConfigSectionTemplate
+                  sectionKey="lpr"
+                  level="replay"
+                  cameraName={replayCamera}
+                  skipSave
+                  noStickyButtons
+                  requiresRestart={false}
+                  collapsible
+                  defaultCollapsed={false}
+                  showTitle
+                  showOverrideIndicator={false}
+                />
+              )}
+            </div>
+          )}
+        </>
+      }
+      open={open}
+      onOpenChange={setOpen}
+    />
+  );
+}

web/src/components/overlay/detail/DetailActionsMenu.tsx

@@ -63,8 +63,8 @@ export default function DetailActionsMenu({
     axios
       .post("debug_replay/start", {
         camera: search.camera,
-        start_time: search.start_time,
-        end_time: search.end_time,
+        start_time: (search.start_time ?? 0) - REVIEW_PADDING,
+        end_time: (search.end_time ?? Date.now() / 1000) + REVIEW_PADDING,
       })
       .then((response) => {
         if (response.status === 202 || response.status === 200) {

web/src/components/timeline/EventMenu.tsx

@@ -12,6 +12,7 @@ import { baseUrl } from "@/api/baseUrl";
 import { useNavigate } from "react-router-dom";
 import { useTranslation } from "react-i18next";
 import { Event } from "@/types/event";
+import { REVIEW_PADDING } from "@/types/review";
 import { FrigateConfig } from "@/types/frigateConfig";
 import { useCallback, useState } from "react";
 import { useIsAdmin } from "@/hooks/use-is-admin";
@@ -58,8 +59,8 @@ export default function EventMenu({
       axios
         .post("debug_replay/start", {
           camera: event.camera,
-          start_time: event.start_time,
-          end_time: event.end_time,
+          start_time: (event.start_time ?? 0) - REVIEW_PADDING,
+          end_time: (event.end_time ?? Date.now() / 1000) + REVIEW_PADDING,
         })
         .then((response) => {
           if (response.status === 202 || response.status === 200) {

web/src/pages/Replay.tsx

@@ -27,7 +27,7 @@ import {
   PopoverContent,
   PopoverTrigger,
 } from "@/components/ui/popover";
-import { PlatformAwareSheet } from "@/components/overlay/dialog/PlatformAwareDialog";
+import { DebugReplayConfigSheet } from "@/components/overlay/DebugReplayConfigSheet";
 import { useCameraActivity } from "@/hooks/use-camera-activity";
 import { cn } from "@/lib/utils";
 import Heading from "@/components/ui/heading";
@@ -40,16 +40,14 @@ import { Progress } from "@/components/ui/progress";
 import { ObjectType } from "@/types/ws";
 import { useJobStatus } from "@/api/ws";
 import WsMessageFeed from "@/components/ws/WsMessageFeed";
-import { ConfigSectionTemplate } from "@/components/config-form/sections/ConfigSectionTemplate";
 
-import { LuExternalLink, LuInfo, LuSettings } from "react-icons/lu";
+import { LuExternalLink, LuInfo } from "react-icons/lu";
 import { LuSquare } from "react-icons/lu";
 import { MdReplay } from "react-icons/md";
 import { isDesktop, isMobile } from "react-device-detect";
 import Logo from "@/components/Logo";
 import { Separator } from "@/components/ui/separator";
 import { useDocDomain } from "@/hooks/use-doc-domain";
-import { useConfigSchema } from "@/hooks/use-config-schema";
 import DebugDrawingLayer from "@/components/overlay/DebugDrawingLayer";
 import { IoMdArrowRoundBack } from "react-icons/io";
 
@@ -125,7 +123,6 @@ export default function Replay() {
   });
   const { payload: replayJob } =
     useJobStatus<DebugReplayJobResults>("debug_replay");
-  const configSchema = useConfigSchema();
   const [isInitializing, setIsInitializing] = useState(true);
 
   // Refresh status immediately on mount to avoid showing "no session" briefly
@@ -139,7 +136,6 @@ export default function Replay() {
 
   const [options, setOptions] = useState<DebugOptions>(DEFAULT_OPTIONS);
   const [isStopping, setIsStopping] = useState(false);
-  const [configDialogOpen, setConfigDialogOpen] = useState(false);
 
   const searchParams = useMemo(() => {
     const params = new URLSearchParams();
@@ -327,91 +323,8 @@ export default function Replay() {
           )}
         </Button>
         <div className="flex items-center gap-2">
-          <PlatformAwareSheet
-            trigger={
-              <Button
-                variant="outline"
-                size="sm"
-                className="flex items-center gap-2"
-              >
-                <LuSettings className="size-4" />
-                <span className="hidden md:inline">
-                  {t("page.configuration")}
-                </span>
-              </Button>
-            }
-            title={t("page.configuration")}
-            titleClassName="text-lg font-semibold"
-            contentClassName="scrollbar-container flex flex-col gap-0 overflow-y-auto px-6 pb-6 sm:max-w-xl md:max-w-2xl xl:max-w-3xl"
-            content={
-              <>
-                <p className="mb-5 text-sm text-muted-foreground">
-                  {t("page.configurationDesc")}
-                </p>
-                {configSchema == null ? (
-                  <div className="flex h-40 items-center justify-center">
-                    <ActivityIndicator />
-                  </div>
-                ) : (
-                  <div className="space-y-6">
-                    <ConfigSectionTemplate
-                      sectionKey="motion"
-                      level="replay"
-                      cameraName={status.replay_camera ?? undefined}
-                      skipSave
-                      noStickyButtons
-                      requiresRestart={false}
-                      collapsible
-                      defaultCollapsed={false}
-                      showTitle
-                      showOverrideIndicator={false}
-                    />
-                    <ConfigSectionTemplate
-                      sectionKey="objects"
-                      level="replay"
-                      cameraName={status.replay_camera ?? undefined}
-                      skipSave
-                      noStickyButtons
-                      requiresRestart={false}
-                      collapsible
-                      defaultCollapsed={false}
-                      showTitle
-                      showOverrideIndicator={false}
-                    />
-                    {config?.face_recognition?.enabled && (
-                      <ConfigSectionTemplate
-                        sectionKey="face_recognition"
-                        level="replay"
-                        cameraName={status.replay_camera ?? undefined}
-                        skipSave
-                        noStickyButtons
-                        requiresRestart={false}
-                        collapsible
-                        defaultCollapsed={false}
-                        showTitle
-                        showOverrideIndicator={false}
-                      />
-                    )}
-                    {config?.lpr?.enabled && (
-                      <ConfigSectionTemplate
-                        sectionKey="lpr"
-                        level="replay"
-                        cameraName={status.replay_camera ?? undefined}
-                        skipSave
-                        noStickyButtons
-                        requiresRestart={false}
-                        collapsible
-                        defaultCollapsed={false}
-                        showTitle
-                        showOverrideIndicator={false}
-                      />
-                    )}
-                  </div>
-                )}
-              </>
-            }
-            open={configDialogOpen}
-            onOpenChange={setConfigDialogOpen}
+          <DebugReplayConfigSheet
+            replayCamera={status.replay_camera ?? undefined}
           />
 
           <AlertDialog>

web/src/utils/configUtil.ts

@@ -229,7 +229,12 @@ export function buildOverrides(
 
     const result: JsonObject = {};
     for (const [key, value] of Object.entries(currentObj)) {
-      if (value === undefined && baseObj && baseObj[key] !== undefined) {
+      if (
+        (value === undefined || value === null) &&
+        baseObj &&
+        baseObj[key] !== undefined &&
+        baseObj[key] !== null
+      ) {
         result[key] = "";
         continue;
       }

web/src/views/settings/CameraManagementView.tsx

@@ -36,7 +36,15 @@ import axios from "axios";
 import ActivityIndicator from "@/components/indicators/activity-indicator";
 import RestartDialog from "@/components/overlay/dialog/RestartDialog";
 import RestartRequiredIndicator from "@/components/indicators/RestartRequiredIndicator";
-import TextEntryDialog from "@/components/overlay/dialog/TextEntryDialog";
+import {
+  Dialog,
+  DialogContent,
+  DialogDescription,
+  DialogFooter,
+  DialogHeader,
+  DialogTitle,
+} from "@/components/ui/dialog";
+import { Input } from "@/components/ui/input";
 import {
   Tooltip,
   TooltipContent,
@@ -53,6 +61,17 @@ import {
   SelectTrigger,
   SelectValue,
 } from "@/components/ui/select";
+import {
+  Form,
+  FormControl,
+  FormField,
+  FormItem,
+  FormLabel,
+  FormMessage,
+} from "@/components/ui/form";
+import { useForm } from "react-hook-form";
+import { zodResolver } from "@hookform/resolvers/zod";
+import { z } from "zod";
 
 const REORDER_SAVED_INDICATOR_MS = 1500;
 
@@ -482,7 +501,7 @@ function EnabledCameraRow({
           <LuGripVertical className="size-4" />
         </button>
         <CameraNameLabel camera={camera} />
-        <CameraFriendlyNameEditor
+        <CameraDetailsEditor
           cameraName={camera}
           onConfigChanged={onConfigChanged}
         />
@@ -519,25 +538,91 @@ function CameraEnableSwitch({ cameraName }: CameraEnableSwitchProps) {
   );
 }
 
-type CameraFriendlyNameEditorProps = {
+type CameraDetailsEditorProps = {
   cameraName: string;
   onConfigChanged: () => Promise<unknown>;
 };
 
-function CameraFriendlyNameEditor({
+type CameraDetailsFormValues = {
+  friendlyName: string;
+  webuiUrl: string;
+};
+
+function CameraDetailsEditor({
   cameraName,
   onConfigChanged,
-}: CameraFriendlyNameEditorProps) {
+}: CameraDetailsEditorProps) {
   const { t } = useTranslation(["views/settings", "common"]);
   const { data: config } = useSWR<FrigateConfig>("config");
   const [open, setOpen] = useState(false);
   const [isSaving, setIsSaving] = useState(false);
 
   const currentFriendlyName = config?.cameras?.[cameraName]?.friendly_name;
+  const currentWebuiUrl = config?.cameras?.[cameraName]?.webui_url;
 
-  const onSave = useCallback(
-    async (text: string) => {
+  const formSchema = useMemo(
+    () =>
+      z.object({
+        friendlyName: z.string(),
+        webuiUrl: z.string().refine(
+          (val) => {
+            const trimmed = val.trim();
+            if (!trimmed) return true;
+            try {
+              new URL(trimmed);
+              return true;
+            } catch {
+              return false;
+            }
+          },
+          {
+            message: t("cameraManagement.streams.details.webuiUrlInvalid", {
+              ns: "views/settings",
+            }),
+          },
+        ),
+      }),
+    [t],
+  );
+
+  const form = useForm<CameraDetailsFormValues>({
+    resolver: zodResolver(formSchema),
+    defaultValues: {
+      friendlyName: currentFriendlyName ?? "",
+      webuiUrl: currentWebuiUrl ?? "",
+    },
+  });
+
+  // Reset form values from config whenever the dialog is opened.
+  useEffect(() => {
+    if (open) {
+      form.reset({
+        friendlyName: currentFriendlyName ?? "",
+        webuiUrl: currentWebuiUrl ?? "",
+      });
+    }
+  }, [open, currentFriendlyName, currentWebuiUrl, form]);
+
+  const onSubmit = useCallback(
+    async (values: CameraDetailsFormValues) => {
       if (isSaving) return;
+
+      // only send fields the user actually changed
+      const newFriendly = values.friendlyName.trim() || null;
+      const newWebui = values.webuiUrl.trim() || null;
+      const cameraUpdate: Record<string, string | null> = {};
+      if (newFriendly !== (currentFriendlyName ?? null)) {
+        cameraUpdate.friendly_name = newFriendly;
+      }
+      if (newWebui !== (currentWebuiUrl ?? null)) {
+        cameraUpdate.webui_url = newWebui;
+      }
+
+      if (Object.keys(cameraUpdate).length === 0) {
+        setOpen(false);
+        return;
+      }
+
       setIsSaving(true);
 
       try {
@@ -545,9 +630,7 @@ function CameraFriendlyNameEditor({
           requires_restart: 0,
           config_data: {
             cameras: {
-              [cameraName]: {
-                friendly_name: text.trim() || null,
-              },
+              [cameraName]: cameraUpdate,
             },
           },
         });
@@ -573,10 +656,17 @@ function CameraFriendlyNameEditor({
         setIsSaving(false);
       }
     },
-    [cameraName, isSaving, onConfigChanged, t],
+    [
+      cameraName,
+      currentFriendlyName,
+      currentWebuiUrl,
+      isSaving,
+      onConfigChanged,
+      t,
+    ],
   );
 
-  const renameLabel = t("cameraManagement.streams.friendlyName.rename", {
+  const editLabel = t("cameraManagement.streams.details.edit", {
     ns: "views/settings",
   });
 
@@ -588,30 +678,107 @@ function CameraFriendlyNameEditor({
             variant="ghost"
             size="icon"
             className="size-7"
-            aria-label={renameLabel}
+            aria-label={editLabel}
             onClick={() => setOpen(true)}
             disabled={isSaving}
           >
             <LuPencil className="size-3.5" />
           </Button>
         </TooltipTrigger>
-        <TooltipContent>{renameLabel}</TooltipContent>
+        <TooltipContent>{editLabel}</TooltipContent>
       </Tooltip>
-      <TextEntryDialog
-        open={open}
-        setOpen={setOpen}
-        title={t("cameraManagement.streams.friendlyName.title", {
-          ns: "views/settings",
-        })}
-        description={t("cameraManagement.streams.friendlyName.description", {
-          ns: "views/settings",
-        })}
-        defaultValue={currentFriendlyName ?? ""}
-        placeholder={currentFriendlyName ? undefined : cameraName}
-        allowEmpty
-        isSaving={isSaving}
-        onSave={onSave}
-      />
+      <Dialog open={open} onOpenChange={setOpen}>
+        <DialogContent>
+          <DialogHeader>
+            <DialogTitle>
+              {t("cameraManagement.streams.details.title", {
+                ns: "views/settings",
+              })}
+            </DialogTitle>
+            <DialogDescription>
+              {t("cameraManagement.streams.details.description", {
+                ns: "views/settings",
+              })}
+            </DialogDescription>
+          </DialogHeader>
+          <Form {...form}>
+            <form onSubmit={form.handleSubmit(onSubmit)} className="space-y-4">
+              <FormField
+                control={form.control}
+                name="friendlyName"
+                render={({ field }) => (
+                  <FormItem>
+                    <FormLabel>
+                      {t("cameraManagement.streams.details.friendlyNameLabel", {
+                        ns: "views/settings",
+                      })}
+                    </FormLabel>
+                    <FormControl>
+                      <Input
+                        {...field}
+                        placeholder={cameraName}
+                        disabled={isSaving}
+                        autoFocus
+                      />
+                    </FormControl>
+                    <p className="text-xs text-muted-foreground">
+                      {t("cameraManagement.streams.details.friendlyNameHelp", {
+                        ns: "views/settings",
+                      })}
+                    </p>
+                    <FormMessage />
+                  </FormItem>
+                )}
+              />
+              <FormField
+                control={form.control}
+                name="webuiUrl"
+                render={({ field }) => (
+                  <FormItem>
+                    <FormLabel>
+                      {t("cameraManagement.streams.details.webuiUrlLabel", {
+                        ns: "views/settings",
+                      })}
+                    </FormLabel>
+                    <FormControl>
+                      <Input
+                        {...field}
+                        placeholder="https://"
+                        disabled={isSaving}
+                      />
+                    </FormControl>
+                    <p className="text-xs text-muted-foreground">
+                      {t("cameraManagement.streams.details.webuiUrlHelp", {
+                        ns: "views/settings",
+                      })}
+                    </p>
+                    <FormMessage />
+                  </FormItem>
+                )}
+              />
+              <DialogFooter className="pt-2">
+                <Button
+                  type="button"
+                  disabled={isSaving}
+                  onClick={() => setOpen(false)}
+                >
+                  {t("button.cancel", { ns: "common" })}
+                </Button>
+                <Button variant="select" type="submit" disabled={isSaving}>
+                  {isSaving ? (
+                    <div className="flex flex-row items-center gap-2">
+                      <ActivityIndicator className="size-4" />
+                      <span>{t("button.saving", { ns: "common" })}</span>
+                    </div>
+                  ) : (
+                    t("button.save", { ns: "common" })
+                  )}
+                </Button>
+              </DialogFooter>
+            </form>
+          </Form>
+        </DialogContent>
+      </Dialog>
     </>
   );
 }