diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 2b04d046721b..95e7619a195e 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -56,7 +56,7 @@ jobs:
       - name: Check out code from GitHub
         uses: actions/checkout@v3.5.2
       - name: Run yamllint
-        uses: frenck/action-yamllint@v1.4.0
+        uses: frenck/action-yamllint@v1.4.1
 
   black:
     name: Check black
diff --git a/docker/Dockerfile b/docker/Dockerfile
index 95b6677815dc..2d9a8a9ae487 100644
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -29,6 +29,8 @@ RUN \
         git=1:2.30.2-1+deb11u2 \
         curl=7.74.0-1.3+deb11u7 \
         openssh-client=1:8.4p1-5+deb11u1 \
+        libcairo2=1.16.0-5 \
+        python3-cffi=1.14.5-1 \
     && rm -rf \
         /tmp/* \
         /var/{cache,log}/* \
diff --git a/esphome/__main__.py b/esphome/__main__.py
index 603a06bdd239..c7c83ad83b2f 100644
--- a/esphome/__main__.py
+++ b/esphome/__main__.py
@@ -980,7 +980,7 @@ def run_esphome(argv):
             _LOGGER.error(e, exc_info=args.verbose)
             return 1
 
-    safe_print(f"ESPHome {const.__version__}")
+    _LOGGER.info("ESPHome %s", const.__version__)
 
     for conf_path in args.configuration:
         if any(os.path.basename(conf_path) == x for x in SECRETS_FILES):
diff --git a/esphome/components/api/api.proto b/esphome/components/api/api.proto
index 4cc98c91d91a..34678fde0f4d 100644
--- a/esphome/components/api/api.proto
+++ b/esphome/components/api/api.proto
@@ -1397,6 +1397,7 @@ message VoiceAssistantRequest {
   option (ifdef) = "USE_VOICE_ASSISTANT";
 
   bool start = 1;
+  string conversation_id = 2;
 }
 
 message VoiceAssistantResponse {
diff --git a/esphome/components/api/api_connection.cpp b/esphome/components/api/api_connection.cpp
index c350197e68c5..3983c6403b8d 100644
--- a/esphome/components/api/api_connection.cpp
+++ b/esphome/components/api/api_connection.cpp
@@ -895,11 +895,12 @@ BluetoothConnectionsFreeResponse APIConnection::subscribe_bluetooth_connections_
 #endif
 
 #ifdef USE_VOICE_ASSISTANT
-bool APIConnection::request_voice_assistant(bool start) {
+bool APIConnection::request_voice_assistant(bool start, const std::string &conversation_id) {
   if (!this->voice_assistant_subscription_)
     return false;
   VoiceAssistantRequest msg;
   msg.start = start;
+  msg.conversation_id = conversation_id;
   return this->send_voice_assistant_request(msg);
 }
 void APIConnection::on_voice_assistant_response(const VoiceAssistantResponse &msg) {
diff --git a/esphome/components/api/api_connection.h b/esphome/components/api/api_connection.h
index 78ecbb98e65b..d4e9cc656e4a 100644
--- a/esphome/components/api/api_connection.h
+++ b/esphome/components/api/api_connection.h
@@ -128,7 +128,7 @@ class APIConnection : public APIServerConnection {
   void subscribe_voice_assistant(const SubscribeVoiceAssistantRequest &msg) override {
     this->voice_assistant_subscription_ = msg.subscribe;
   }
-  bool request_voice_assistant(bool start);
+  bool request_voice_assistant(bool start, const std::string &conversation_id);
   void on_voice_assistant_response(const VoiceAssistantResponse &msg) override;
   void on_voice_assistant_event_response(const VoiceAssistantEventResponse &msg) override;
 #endif
diff --git a/esphome/components/api/api_pb2.cpp b/esphome/components/api/api_pb2.cpp
index 1dd8c82e00fe..29e8e207ddd4 100644
--- a/esphome/components/api/api_pb2.cpp
+++ b/esphome/components/api/api_pb2.cpp
@@ -6187,7 +6187,20 @@ bool VoiceAssistantRequest::decode_varint(uint32_t field_id, ProtoVarInt value)
       return false;
   }
 }
-void VoiceAssistantRequest::encode(ProtoWriteBuffer buffer) const { buffer.encode_bool(1, this->start); }
+bool VoiceAssistantRequest::decode_length(uint32_t field_id, ProtoLengthDelimited value) {
+  switch (field_id) {
+    case 2: {
+      this->conversation_id = value.as_string();
+      return true;
+    }
+    default:
+      return false;
+  }
+}
+void VoiceAssistantRequest::encode(ProtoWriteBuffer buffer) const {
+  buffer.encode_bool(1, this->start);
+  buffer.encode_string(2, this->conversation_id);
+}
 #ifdef HAS_PROTO_MESSAGE_DUMP
 void VoiceAssistantRequest::dump_to(std::string &out) const {
   __attribute__((unused)) char buffer[64];
@@ -6195,6 +6208,10 @@ void VoiceAssistantRequest::dump_to(std::string &out) const {
   out.append("  start: ");
   out.append(YESNO(this->start));
   out.append("\n");
+
+  out.append("  conversation_id: ");
+  out.append("'").append(this->conversation_id).append("'");
+  out.append("\n");
   out.append("}");
 }
 #endif
diff --git a/esphome/components/api/api_pb2.h b/esphome/components/api/api_pb2.h
index 0f4b79de1978..cd1cfb595a33 100644
--- a/esphome/components/api/api_pb2.h
+++ b/esphome/components/api/api_pb2.h
@@ -1604,12 +1604,14 @@ class SubscribeVoiceAssistantRequest : public ProtoMessage {
 class VoiceAssistantRequest : public ProtoMessage {
  public:
   bool start{false};
+  std::string conversation_id{};
   void encode(ProtoWriteBuffer buffer) const override;
 #ifdef HAS_PROTO_MESSAGE_DUMP
   void dump_to(std::string &out) const override;
 #endif
 
  protected:
+  bool decode_length(uint32_t field_id, ProtoLengthDelimited value) override;
   bool decode_varint(uint32_t field_id, ProtoVarInt value) override;
 };
 class VoiceAssistantResponse : public ProtoMessage {
diff --git a/esphome/components/api/api_server.cpp b/esphome/components/api/api_server.cpp
index 068f74315c08..3dd47c4dd8d7 100644
--- a/esphome/components/api/api_server.cpp
+++ b/esphome/components/api/api_server.cpp
@@ -428,16 +428,16 @@ void APIServer::on_shutdown() {
 }
 
 #ifdef USE_VOICE_ASSISTANT
-bool APIServer::start_voice_assistant() {
+bool APIServer::start_voice_assistant(const std::string &conversation_id) {
   for (auto &c : this->clients_) {
-    if (c->request_voice_assistant(true))
+    if (c->request_voice_assistant(true, conversation_id))
       return true;
   }
   return false;
 }
 void APIServer::stop_voice_assistant() {
   for (auto &c : this->clients_) {
-    if (c->request_voice_assistant(false))
+    if (c->request_voice_assistant(false, ""))
       return;
   }
 }
diff --git a/esphome/components/api/api_server.h b/esphome/components/api/api_server.h
index a1bec2802fc4..79ba7b17f1bd 100644
--- a/esphome/components/api/api_server.h
+++ b/esphome/components/api/api_server.h
@@ -96,7 +96,7 @@ class APIServer : public Component, public Controller {
 #endif
 
 #ifdef USE_VOICE_ASSISTANT
-  bool start_voice_assistant();
+  bool start_voice_assistant(const std::string &conversation_id);
   void stop_voice_assistant();
 #endif
 
diff --git a/esphome/components/esp32/boards.py b/esphome/components/esp32/boards.py
index e4fdaec0aa2f..30297654bc5e 100644
--- a/esphome/components/esp32/boards.py
+++ b/esphome/components/esp32/boards.py
@@ -42,6 +42,39 @@
 }
 
 ESP32_BOARD_PINS = {
+    "adafruit_feather_esp32s2_tft": {
+        "BUTTON": 0,
+        "A0": 18,
+        "A1": 17,
+        "A2": 16,
+        "A3": 15,
+        "A4": 14,
+        "A5": 8,
+        "SCK": 36,
+        "MOSI": 35,
+        "MISO": 37,
+        "RX": 2,
+        "TX": 1,
+        "D13": 13,
+        "D12": 12,
+        "D11": 11,
+        "D10": 10,
+        "D9": 9,
+        "D6": 6,
+        "D5": 5,
+        "NEOPIXEL": 33,
+        "PIN_NEOPIXEL": 33,
+        "NEOPIXEL_POWER": 34,
+        "SCL": 41,
+        "SDA": 42,
+        "TFT_I2C_POWER": 21,
+        "TFT_CS": 7,
+        "TFT_DC": 39,
+        "TFT_RESET": 40,
+        "TFT_BACKLIGHT": 45,
+        "LED": 13,
+        "LED_BUILTIN": 13,
+    },
     "adafruit_qtpy_esp32c3": {
         "A0": 4,
         "A1": 3,
diff --git a/esphome/components/esp32_rmt_led_strip/led_strip.cpp b/esphome/components/esp32_rmt_led_strip/led_strip.cpp
index eec1bdc992f5..df6ee2ce2f2f 100644
--- a/esphome/components/esp32_rmt_led_strip/led_strip.cpp
+++ b/esphome/components/esp32_rmt_led_strip/led_strip.cpp
@@ -1,3 +1,4 @@
+#include <cinttypes>
 #include "led_strip.h"
 
 #ifdef USE_ESP32
@@ -195,7 +196,7 @@ void ESP32RMTLEDStripLightOutput::dump_config() {
       break;
   }
   ESP_LOGCONFIG(TAG, "  RGB Order: %s", rgb_order);
-  ESP_LOGCONFIG(TAG, "  Max refresh rate: %u", *this->max_refresh_rate_);
+  ESP_LOGCONFIG(TAG, "  Max refresh rate: %" PRIu32, *this->max_refresh_rate_);
   ESP_LOGCONFIG(TAG, "  Number of LEDs: %u", this->num_leds_);
 }
 
diff --git a/esphome/components/i2s_audio/microphone/__init__.py b/esphome/components/i2s_audio/microphone/__init__.py
index 07f51581887a..b917da30450a 100644
--- a/esphome/components/i2s_audio/microphone/__init__.py
+++ b/esphome/components/i2s_audio/microphone/__init__.py
@@ -62,7 +62,7 @@ def validate_esp32_variant(config):
         cv.GenerateID(): cv.declare_id(I2SAudioMicrophone),
         cv.GenerateID(CONF_I2S_AUDIO_ID): cv.use_id(I2SAudioComponent),
         cv.Optional(CONF_CHANNEL, default="right"): cv.enum(CHANNELS),
-        cv.Optional(CONF_BITS_PER_SAMPLE, default="16bit"): cv.All(
+        cv.Optional(CONF_BITS_PER_SAMPLE, default="32bit"): cv.All(
             _validate_bits, cv.enum(BITS_PER_SAMPLE)
         ),
     }
diff --git a/esphome/components/image/__init__.py b/esphome/components/image/__init__.py
index 4260636fa999..113c5a2df1c2 100644
--- a/esphome/components/image/__init__.py
+++ b/esphome/components/image/__init__.py
@@ -1,5 +1,10 @@
 import logging
 
+import io
+from pathlib import Path
+import re
+import requests
+
 from esphome import core
 from esphome.components import display, font
 import esphome.config_validation as cv
@@ -7,15 +12,19 @@
 from esphome.const import (
     CONF_DITHER,
     CONF_FILE,
+    CONF_ICON,
     CONF_ID,
+    CONF_PATH,
     CONF_RAW_DATA_ID,
     CONF_RESIZE,
+    CONF_SOURCE,
     CONF_TYPE,
 )
 from esphome.core import CORE, HexInt
 
 _LOGGER = logging.getLogger(__name__)
 
+DOMAIN = "image"
 DEPENDENCIES = ["display"]
 MULTI_CONF = True
 
@@ -31,9 +40,58 @@
 
 CONF_USE_TRANSPARENCY = "use_transparency"
 
+# If the MDI file cannot be downloaded within this time, abort.
+MDI_DOWNLOAD_TIMEOUT = 30  # seconds
+
+SOURCE_LOCAL = "local"
+SOURCE_MDI = "mdi"
+
 Image_ = display.display_ns.class_("Image")
 
 
+def _compute_local_icon_path(value) -> Path:
+    base_dir = Path(CORE.config_dir) / ".esphome" / DOMAIN / "mdi"
+    return base_dir / f"{value[CONF_ICON]}.svg"
+
+
+def download_mdi(value):
+    mdi_id = value[CONF_ICON]
+    path = _compute_local_icon_path(value)
+    if path.is_file():
+        return value
+    url = f"https://raw.githubusercontent.com/Templarian/MaterialDesign/master/svg/{mdi_id}.svg"
+    _LOGGER.debug("Downloading %s MDI image from %s", mdi_id, url)
+    try:
+        req = requests.get(url, timeout=MDI_DOWNLOAD_TIMEOUT)
+        req.raise_for_status()
+    except requests.exceptions.RequestException as e:
+        raise cv.Invalid(f"Could not download MDI image {mdi_id} from {url}: {e}")
+
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_bytes(req.content)
+    return value
+
+
+def validate_cairosvg_installed(value):
+    """Validate that cairosvg is installed"""
+    try:
+        import cairosvg
+    except ImportError as err:
+        raise cv.Invalid(
+            "Please install the cairosvg python package to use this feature. "
+            "(pip install cairosvg)"
+        ) from err
+
+    major, minor, _ = cairosvg.__version__.split(".")
+    if major < "2" or major == "2" and minor < "2":
+        raise cv.Invalid(
+            "Please update your cairosvg installation to at least 2.2.0. "
+            "(pip install -U cairosvg)"
+        )
+
+    return value
+
+
 def validate_cross_dependencies(config):
     """
     Validate fields whose possible values depend on other fields.
@@ -41,6 +99,13 @@ def validate_cross_dependencies(config):
     have "use_transparency" set to True.
     Also set the default value for those kind of dependent fields.
     """
+    is_mdi = CONF_FILE in config and config[CONF_FILE][CONF_SOURCE] == SOURCE_MDI
+    if CONF_TYPE not in config:
+        if is_mdi:
+            config[CONF_TYPE] = "TRANSPARENT_BINARY"
+        else:
+            config[CONF_TYPE] = "BINARY"
+
     image_type = config[CONF_TYPE]
     is_transparent_type = image_type in ["TRANSPARENT_BINARY", "RGBA"]
 
@@ -51,16 +116,74 @@ def validate_cross_dependencies(config):
     if is_transparent_type and not config[CONF_USE_TRANSPARENCY]:
         raise cv.Invalid(f"Image type {image_type} must always be transparent.")
 
+    if is_mdi and config[CONF_TYPE] not in ["BINARY", "TRANSPARENT_BINARY"]:
+        raise cv.Invalid("MDI images must be binary images.")
+
     return config
 
 
+def validate_file_shorthand(value):
+    value = cv.string_strict(value)
+    if value.startswith("mdi:"):
+        validate_cairosvg_installed(value)
+
+        match = re.search(r"mdi:([a-zA-Z0-9\-]+)", value)
+        if match is None:
+            raise cv.Invalid("Could not parse mdi icon name.")
+        icon = match.group(1)
+        return FILE_SCHEMA(
+            {
+                CONF_SOURCE: SOURCE_MDI,
+                CONF_ICON: icon,
+            }
+        )
+    return FILE_SCHEMA(
+        {
+            CONF_SOURCE: SOURCE_LOCAL,
+            CONF_PATH: value,
+        }
+    )
+
+
+LOCAL_SCHEMA = cv.Schema(
+    {
+        cv.Required(CONF_PATH): cv.file_,
+    }
+)
+
+MDI_SCHEMA = cv.All(
+    {
+        cv.Required(CONF_ICON): cv.string,
+    },
+    download_mdi,
+)
+
+TYPED_FILE_SCHEMA = cv.typed_schema(
+    {
+        SOURCE_LOCAL: LOCAL_SCHEMA,
+        SOURCE_MDI: MDI_SCHEMA,
+    },
+    key=CONF_SOURCE,
+)
+
+
+def _file_schema(value):
+    if isinstance(value, str):
+        return validate_file_shorthand(value)
+    return TYPED_FILE_SCHEMA(value)
+
+
+FILE_SCHEMA = cv.Schema(_file_schema)
+
 IMAGE_SCHEMA = cv.Schema(
     cv.All(
         {
             cv.Required(CONF_ID): cv.declare_id(Image_),
-            cv.Required(CONF_FILE): cv.file_,
+            cv.Required(CONF_FILE): FILE_SCHEMA,
             cv.Optional(CONF_RESIZE): cv.dimensions,
-            cv.Optional(CONF_TYPE, default="BINARY"): cv.enum(IMAGE_TYPE, upper=True),
+            # Not setting default here on purpose; the default depends on the source type
+            # (file or mdi), and will be set in the "validate_cross_dependencies" validator.
+            cv.Optional(CONF_TYPE): cv.enum(IMAGE_TYPE, upper=True),
             # Not setting default here on purpose; the default depends on the image type,
             # and thus will be set in the "validate_cross_dependencies" validator.
             cv.Optional(CONF_USE_TRANSPARENCY): cv.boolean,
@@ -79,24 +202,43 @@ def validate_cross_dependencies(config):
 async def to_code(config):
     from PIL import Image
 
-    path = CORE.relative_config_path(config[CONF_FILE])
-    try:
-        image = Image.open(path)
-    except Exception as e:
-        raise core.EsphomeError(f"Could not load image file {path}: {e}")
+    conf_file = config[CONF_FILE]
 
-    width, height = image.size
+    if conf_file[CONF_SOURCE] == SOURCE_LOCAL:
+        path = CORE.relative_config_path(conf_file[CONF_PATH])
+        try:
+            image = Image.open(path)
+        except Exception as e:
+            raise core.EsphomeError(f"Could not load image file {path}: {e}")
+        if CONF_RESIZE in config:
+            image.thumbnail(config[CONF_RESIZE])
+    elif conf_file[CONF_SOURCE] == SOURCE_MDI:
+        # Those imports are only needed in case of MDI images; adding them
+        # to the top would force configurations not using MDI to also have them
+        # installed for no reason.
+        from cairosvg import svg2png
 
-    if CONF_RESIZE in config:
-        image.thumbnail(config[CONF_RESIZE])
-        width, height = image.size
-    else:
-        if width > 500 or height > 500:
-            _LOGGER.warning(
-                'The image "%s" you requested is very big. Please consider'
-                " using the resize parameter.",
-                path,
+        svg_file = _compute_local_icon_path(conf_file)
+        if CONF_RESIZE in config:
+            req_width, req_height = config[CONF_RESIZE]
+            svg_image = svg2png(
+                url=svg_file.as_posix(),
+                output_width=req_width,
+                output_height=req_height,
             )
+        else:
+            svg_image = svg2png(url=svg_file.as_posix())
+
+        image = Image.open(io.BytesIO(svg_image))
+
+    width, height = image.size
+
+    if CONF_RESIZE not in config and (width > 500 or height > 500):
+        _LOGGER.warning(
+            'The image "%s" you requested is very big. Please consider'
+            " using the resize parameter.",
+            path,
+        )
 
     transparent = config[CONF_USE_TRANSPARENCY]
 
diff --git a/esphome/components/light/light_call.cpp b/esphome/components/light/light_call.cpp
index fb4e45b3b6c5..8dc5d4fbe7f6 100644
--- a/esphome/components/light/light_call.cpp
+++ b/esphome/components/light/light_call.cpp
@@ -1,3 +1,4 @@
+#include <cinttypes>
 #include "light_call.h"
 #include "light_state.h"
 #include "esphome/core/log.h"
@@ -283,7 +284,7 @@ LightColorValues LightCall::validate_() {
 
   // validate effect index
   if (this->has_effect_() && *this->effect_ > this->parent_->effects_.size()) {
-    ESP_LOGW(TAG, "'%s' - Invalid effect index %u!", name, *this->effect_);
+    ESP_LOGW(TAG, "'%s' - Invalid effect index %" PRIu32 "!", name, *this->effect_);
     this->effect_.reset();
   }
 
diff --git a/esphome/components/ota/ota_backend_esp_idf.cpp b/esphome/components/ota/ota_backend_esp_idf.cpp
index 7688629e396b..319a1482f16c 100644
--- a/esphome/components/ota/ota_backend_esp_idf.cpp
+++ b/esphome/components/ota/ota_backend_esp_idf.cpp
@@ -21,25 +21,35 @@ OTAResponseTypes IDFOTABackend::begin(size_t image_size) {
     return OTA_RESPONSE_ERROR_NO_UPDATE_PARTITION;
   }
 
+#if CONFIG_ESP_TASK_WDT_TIMEOUT_S < 15
   // The following function takes longer than the 5 seconds timeout of WDT
 #if ESP_IDF_VERSION_MAJOR >= 5
   esp_task_wdt_config_t wdtc;
-  wdtc.timeout_ms = 15000;
   wdtc.idle_core_mask = 0;
+#if CONFIG_ESP_TASK_WDT_CHECK_IDLE_TASK_CPU0
+  wdtc.idle_core_mask |= (1 << 0);
+#endif
+#if CONFIG_ESP_TASK_WDT_CHECK_IDLE_TASK_CPU1
+  wdtc.idle_core_mask |= (1 << 1);
+#endif
+  wdtc.timeout_ms = 15000;
   wdtc.trigger_panic = false;
   esp_task_wdt_reconfigure(&wdtc);
 #else
   esp_task_wdt_init(15, false);
+#endif
 #endif
 
   esp_err_t err = esp_ota_begin(this->partition_, image_size, &this->update_handle_);
 
+#if CONFIG_ESP_TASK_WDT_TIMEOUT_S < 15
   // Set the WDT back to the configured timeout
 #if ESP_IDF_VERSION_MAJOR >= 5
-  wdtc.timeout_ms = CONFIG_ESP_TASK_WDT_TIMEOUT_S;
+  wdtc.timeout_ms = CONFIG_ESP_TASK_WDT_TIMEOUT_S * 1000;
   esp_task_wdt_reconfigure(&wdtc);
 #else
   esp_task_wdt_init(CONFIG_ESP_TASK_WDT_TIMEOUT_S, false);
+#endif
 #endif
 
   if (err != ESP_OK) {
diff --git a/esphome/components/ota/ota_component.cpp b/esphome/components/ota/ota_component.cpp
index 39ba3dbed463..acf9e923b625 100644
--- a/esphome/components/ota/ota_component.cpp
+++ b/esphome/components/ota/ota_component.cpp
@@ -99,7 +99,7 @@ void OTAComponent::dump_config() {
 #endif
   if (this->has_safe_mode_ && this->safe_mode_rtc_value_ > 1 &&
       this->safe_mode_rtc_value_ != esphome::ota::OTAComponent::ENTER_SAFE_MODE_MAGIC) {
-    ESP_LOGW(TAG, "Last Boot was an unhandled reset, will proceed to safe mode in %d restarts",
+    ESP_LOGW(TAG, "Last Boot was an unhandled reset, will proceed to safe mode in %" PRIu32 " restarts",
              this->safe_mode_num_attempts_ - this->safe_mode_rtc_value_);
   }
 }
@@ -191,7 +191,7 @@ void OTAComponent::handle_() {
     this->writeall_(buf, 1);
     md5::MD5Digest md5{};
     md5.init();
-    sprintf(sbuf, "%08X", random_uint32());
+    sprintf(sbuf, "%08" PRIx32, random_uint32());
     md5.add(sbuf, 8);
     md5.calculate();
     md5.get_hex(sbuf);
@@ -466,7 +466,7 @@ bool OTAComponent::should_enter_safe_mode(uint8_t num_attempts, uint32_t enable_
   if (is_manual_safe_mode) {
     ESP_LOGI(TAG, "Safe mode has been entered manually");
   } else {
-    ESP_LOGCONFIG(TAG, "There have been %u suspected unsuccessful boot attempts.", this->safe_mode_rtc_value_);
+    ESP_LOGCONFIG(TAG, "There have been %" PRIu32 " suspected unsuccessful boot attempts.", this->safe_mode_rtc_value_);
   }
 
   if (this->safe_mode_rtc_value_ >= num_attempts || is_manual_safe_mode) {
diff --git a/esphome/components/prometheus/prometheus_handler.h b/esphome/components/prometheus/prometheus_handler.h
index f416ecf246fd..0ae2856ce43f 100644
--- a/esphome/components/prometheus/prometheus_handler.h
+++ b/esphome/components/prometheus/prometheus_handler.h
@@ -5,6 +5,7 @@
 #include <map>
 #include <utility>
 
+#include "esphome/core/entity_base.h"
 #include "esphome/components/web_server_base/web_server_base.h"
 #include "esphome/core/controller.h"
 #include "esphome/core/component.h"
diff --git a/esphome/components/rp2040/__init__.py b/esphome/components/rp2040/__init__.py
index 1eba0bf19278..ad66ce6d1803 100644
--- a/esphome/components/rp2040/__init__.py
+++ b/esphome/components/rp2040/__init__.py
@@ -157,7 +157,6 @@ async def to_code(config):
         "platform_packages",
         [
             f"earlephilhower/framework-arduinopico@{conf[CONF_SOURCE]}",
-            "earlephilhower/tool-pioasm-rp2040-earlephilhower",
         ],
     )
 
diff --git a/esphome/components/rp2040_pio_led_strip/light.py b/esphome/components/rp2040_pio_led_strip/light.py
index 432ff6935ab3..a2ba72318f04 100644
--- a/esphome/components/rp2040_pio_led_strip/light.py
+++ b/esphome/components/rp2040_pio_led_strip/light.py
@@ -265,3 +265,9 @@ async def to_code(config):
                 time_to_cycles(config[CONF_BIT1_LOW]),
             ),
         )
+    cg.add_platformio_option(
+        "platform_packages",
+        [
+            "earlephilhower/tool-pioasm-rp2040-earlephilhower",
+        ],
+    )
diff --git a/esphome/components/select/select.h b/esphome/components/select/select.h
index 23258d578500..8ca9a69d1c75 100644
--- a/esphome/components/select/select.h
+++ b/esphome/components/select/select.h
@@ -17,6 +17,13 @@ namespace select {
     } \
   }
 
+#define SUB_SELECT(name) \
+ protected: \
+  select::Select *name##_select_{nullptr}; \
+\
+ public: \
+  void set_##name##_select(select::Select *select) { this->name##_select_ = select; }
+
 /** Base-class for all selects.
  *
  * A select can use publish_state to send out a new value.
diff --git a/esphome/components/sntp/sntp_component.h b/esphome/components/sntp/sntp_component.h
index 4c70a6b09f97..987dd23a19e0 100644
--- a/esphome/components/sntp/sntp_component.h
+++ b/esphome/components/sntp/sntp_component.h
@@ -22,7 +22,7 @@ class SNTPComponent : public time::RealTimeClock {
     this->server_2_ = server_2;
     this->server_3_ = server_3;
   }
-  float get_setup_priority() const override { return setup_priority::AFTER_WIFI; }
+  float get_setup_priority() const override { return setup_priority::BEFORE_CONNECTION; }
 
   void update() override;
   void loop() override;
diff --git a/esphome/components/st7789v/display.py b/esphome/components/st7789v/display.py
index d18e305cc216..a81101f2d1f3 100644
--- a/esphome/components/st7789v/display.py
+++ b/esphome/components/st7789v/display.py
@@ -1,7 +1,7 @@
 import esphome.codegen as cg
 import esphome.config_validation as cv
 from esphome import pins
-from esphome.components import display, spi
+from esphome.components import display, spi, power_supply
 from esphome.const import (
     CONF_BACKLIGHT_PIN,
     CONF_DC_PIN,
@@ -11,6 +11,7 @@
     CONF_MODEL,
     CONF_RESET_PIN,
     CONF_WIDTH,
+    CONF_POWER_SUPPLY,
 )
 from . import st7789v_ns
 
@@ -32,6 +33,7 @@
     "TTGO_TDISPLAY_135X240": ST7789VModel.ST7789V_MODEL_TTGO_TDISPLAY_135_240,
     "ADAFRUIT_FUNHOUSE_240X240": ST7789VModel.ST7789V_MODEL_ADAFRUIT_FUNHOUSE_240_240,
     "ADAFRUIT_RR_280X240": ST7789VModel.ST7789V_MODEL_ADAFRUIT_RR_280_240,
+    "ADAFRUIT_S2_TFT_FEATHER_240X135": ST7789VModel.ST7789V_MODEL_ADAFRUIT_S2_TFT_FEATHER_240_135,
     "CUSTOM": ST7789VModel.ST7789V_MODEL_CUSTOM,
 }
 
@@ -58,6 +60,14 @@ def validate_st7789v(config):
         raise cv.Invalid(
             f'Do not specify {CONF_HEIGHT}, {CONF_WIDTH}, {CONF_OFFSET_HEIGHT} or {CONF_OFFSET_WIDTH} when using {CONF_MODEL} that is not "CUSTOM"'
         )
+
+    if (
+        config[CONF_MODEL].upper() == "ADAFRUIT_S2_TFT_FEATHER_240X135"
+        and CONF_POWER_SUPPLY not in config
+    ):
+        raise cv.Invalid(
+            f'{CONF_POWER_SUPPLY} must be specified when {CONF_MODEL} is "ADAFRUIT_S2_TFT_FEATHER_240X135"'
+        )
     return config
 
 
@@ -69,6 +79,7 @@ def validate_st7789v(config):
             cv.Required(CONF_RESET_PIN): pins.gpio_output_pin_schema,
             cv.Required(CONF_DC_PIN): pins.gpio_output_pin_schema,
             cv.Optional(CONF_BACKLIGHT_PIN): pins.gpio_output_pin_schema,
+            cv.Optional(CONF_POWER_SUPPLY): cv.use_id(power_supply.PowerSupply),
             cv.Optional(CONF_EIGHTBITCOLOR, default=False): cv.boolean,
             cv.Optional(CONF_HEIGHT): cv.int_,
             cv.Optional(CONF_WIDTH): cv.int_,
@@ -113,3 +124,7 @@ async def to_code(config):
             config[CONF_LAMBDA], [(display.DisplayBufferRef, "it")], return_type=cg.void
         )
         cg.add(var.set_writer(lambda_))
+
+    if CONF_POWER_SUPPLY in config:
+        ps = await cg.get_variable(config[CONF_POWER_SUPPLY])
+        cg.add(var.set_power_supply(ps))
diff --git a/esphome/components/st7789v/st7789v.cpp b/esphome/components/st7789v/st7789v.cpp
index 8a4fcfb179e6..0e7c9b9123e8 100644
--- a/esphome/components/st7789v/st7789v.cpp
+++ b/esphome/components/st7789v/st7789v.cpp
@@ -8,6 +8,10 @@ static const char *const TAG = "st7789v";
 
 void ST7789V::setup() {
   ESP_LOGCONFIG(TAG, "Setting up SPI ST7789V...");
+#ifdef USE_POWER_SUPPLY
+  this->power_.request();
+  // the PowerSupply component takes care of post turn-on delay
+#endif
   this->spi_setup();
   this->dc_pin_->setup();  // OUTPUT
 
@@ -128,6 +132,9 @@ void ST7789V::dump_config() {
   LOG_PIN("  Reset Pin: ", this->reset_pin_);
   LOG_PIN("  B/L Pin: ", this->backlight_pin_);
   LOG_UPDATE_INTERVAL(this);
+#ifdef USE_POWER_SUPPLY
+  ESP_LOGCONFIG(TAG, "  Power Supply Configured: yes");
+#endif
 }
 
 float ST7789V::get_setup_priority() const { return setup_priority::PROCESSOR; }
@@ -162,6 +169,13 @@ void ST7789V::set_model(ST7789VModel model) {
       this->offset_width_ = 20;
       break;
 
+    case ST7789V_MODEL_ADAFRUIT_S2_TFT_FEATHER_240_135:
+      this->height_ = 240;
+      this->width_ = 135;
+      this->offset_height_ = 52;
+      this->offset_width_ = 40;
+      break;
+
     default:
       break;
   }
@@ -323,6 +337,8 @@ const char *ST7789V::model_str_() {
       return "Adafruit Funhouse 240x240";
     case ST7789V_MODEL_ADAFRUIT_RR_280_240:
       return "Adafruit Round-Rectangular 280x240";
+    case ST7789V_MODEL_ADAFRUIT_S2_TFT_FEATHER_240_135:
+      return "Adafruit ESP32-S2 TFT Feather";
     default:
       return "Custom";
   }
diff --git a/esphome/components/st7789v/st7789v.h b/esphome/components/st7789v/st7789v.h
index 96e97c9d7865..ccbe50cf8553 100644
--- a/esphome/components/st7789v/st7789v.h
+++ b/esphome/components/st7789v/st7789v.h
@@ -3,6 +3,9 @@
 #include "esphome/core/component.h"
 #include "esphome/components/spi/spi.h"
 #include "esphome/components/display/display_buffer.h"
+#ifdef USE_POWER_SUPPLY
+#include "esphome/components/power_supply/power_supply.h"
+#endif
 
 namespace esphome {
 namespace st7789v {
@@ -11,6 +14,7 @@ enum ST7789VModel {
   ST7789V_MODEL_TTGO_TDISPLAY_135_240,
   ST7789V_MODEL_ADAFRUIT_FUNHOUSE_240_240,
   ST7789V_MODEL_ADAFRUIT_RR_280_240,
+  ST7789V_MODEL_ADAFRUIT_S2_TFT_FEATHER_240_135,
   ST7789V_MODEL_CUSTOM
 };
 
@@ -120,6 +124,9 @@ class ST7789V : public PollingComponent,
   void set_dc_pin(GPIOPin *dc_pin) { this->dc_pin_ = dc_pin; }
   void set_reset_pin(GPIOPin *reset_pin) { this->reset_pin_ = reset_pin; }
   void set_backlight_pin(GPIOPin *backlight_pin) { this->backlight_pin_ = backlight_pin; }
+#ifdef USE_POWER_SUPPLY
+  void set_power_supply(power_supply::PowerSupply *power_supply) { this->power_.set_parent(power_supply); }
+#endif
 
   void set_eightbitcolor(bool eightbitcolor) { this->eightbitcolor_ = eightbitcolor; }
   void set_height(uint32_t height) { this->height_ = height; }
@@ -143,6 +150,9 @@ class ST7789V : public PollingComponent,
   GPIOPin *dc_pin_{nullptr};
   GPIOPin *reset_pin_{nullptr};
   GPIOPin *backlight_pin_{nullptr};
+#ifdef USE_POWER_SUPPLY
+  power_supply::PowerSupplyRequester power_;
+#endif
 
   bool eightbitcolor_{false};
   uint16_t height_{0};
diff --git a/esphome/components/switch/switch.h b/esphome/components/switch/switch.h
index 9daac4ee230e..b5395a2c8397 100644
--- a/esphome/components/switch/switch.h
+++ b/esphome/components/switch/switch.h
@@ -8,6 +8,13 @@
 namespace esphome {
 namespace switch_ {
 
+#define SUB_SWITCH(name) \
+ protected: \
+  switch_::Switch *name##_switch_{nullptr}; \
+\
+ public: \
+  void set_##name##_switch(switch_::Switch *s) { this->name##_switch_ = s; }
+
 // bit0: on/off. bit1: persistent. bit2: inverted. bit3: disabled
 const int RESTORE_MODE_ON_MASK = 0x01;
 const int RESTORE_MODE_PERSISTENT_MASK = 0x02;
diff --git a/esphome/components/voice_assistant/__init__.py b/esphome/components/voice_assistant/__init__.py
index 624fcdf52cba..55d995be88f1 100644
--- a/esphome/components/voice_assistant/__init__.py
+++ b/esphome/components/voice_assistant/__init__.py
@@ -1,16 +1,23 @@
 import esphome.config_validation as cv
 import esphome.codegen as cg
 
-from esphome.const import CONF_ID, CONF_MICROPHONE, CONF_SPEAKER
+from esphome.const import (
+    CONF_ID,
+    CONF_MICROPHONE,
+    CONF_SPEAKER,
+    CONF_MEDIA_PLAYER,
+)
 from esphome import automation
-from esphome.automation import register_action
-from esphome.components import microphone, speaker
+from esphome.automation import register_action, register_condition
+from esphome.components import microphone, speaker, media_player
 
 AUTO_LOAD = ["socket"]
 DEPENDENCIES = ["api", "microphone"]
 
 CODEOWNERS = ["@jesserockz"]
 
+CONF_SILENCE_DETECTION = "silence_detection"
+CONF_ON_LISTENING = "on_listening"
 CONF_ON_START = "on_start"
 CONF_ON_STT_END = "on_stt_end"
 CONF_ON_TTS_START = "on_tts_start"
@@ -25,16 +32,25 @@
 StartAction = voice_assistant_ns.class_(
     "StartAction", automation.Action, cg.Parented.template(VoiceAssistant)
 )
+StartContinuousAction = voice_assistant_ns.class_(
+    "StartContinuousAction", automation.Action, cg.Parented.template(VoiceAssistant)
+)
 StopAction = voice_assistant_ns.class_(
     "StopAction", automation.Action, cg.Parented.template(VoiceAssistant)
 )
+IsRunningCondition = voice_assistant_ns.class_(
+    "IsRunningCondition", automation.Condition, cg.Parented.template(VoiceAssistant)
+)
 
 
 CONFIG_SCHEMA = cv.Schema(
     {
         cv.GenerateID(): cv.declare_id(VoiceAssistant),
         cv.GenerateID(CONF_MICROPHONE): cv.use_id(microphone.Microphone),
-        cv.Optional(CONF_SPEAKER): cv.use_id(speaker.Speaker),
+        cv.Exclusive(CONF_SPEAKER, "output"): cv.use_id(speaker.Speaker),
+        cv.Exclusive(CONF_MEDIA_PLAYER, "output"): cv.use_id(media_player.MediaPlayer),
+        cv.Optional(CONF_SILENCE_DETECTION, default=True): cv.boolean,
+        cv.Optional(CONF_ON_LISTENING): automation.validate_automation(single=True),
         cv.Optional(CONF_ON_START): automation.validate_automation(single=True),
         cv.Optional(CONF_ON_STT_END): automation.validate_automation(single=True),
         cv.Optional(CONF_ON_TTS_START): automation.validate_automation(single=True),
@@ -56,6 +72,17 @@ async def to_code(config):
         spkr = await cg.get_variable(config[CONF_SPEAKER])
         cg.add(var.set_speaker(spkr))
 
+    if CONF_MEDIA_PLAYER in config:
+        mp = await cg.get_variable(config[CONF_MEDIA_PLAYER])
+        cg.add(var.set_media_player(mp))
+
+    cg.add(var.set_silence_detection(config[CONF_SILENCE_DETECTION]))
+
+    if CONF_ON_LISTENING in config:
+        await automation.build_automation(
+            var.get_listening_trigger(), [], config[CONF_ON_LISTENING]
+        )
+
     if CONF_ON_START in config:
         await automation.build_automation(
             var.get_start_trigger(), [], config[CONF_ON_START]
@@ -96,6 +123,11 @@ async def to_code(config):
 VOICE_ASSISTANT_ACTION_SCHEMA = cv.Schema({cv.GenerateID(): cv.use_id(VoiceAssistant)})
 
 
+@register_action(
+    "voice_assistant.start_continuous",
+    StartContinuousAction,
+    VOICE_ASSISTANT_ACTION_SCHEMA,
+)
 @register_action("voice_assistant.start", StartAction, VOICE_ASSISTANT_ACTION_SCHEMA)
 async def voice_assistant_listen_to_code(config, action_id, template_arg, args):
     var = cg.new_Pvariable(action_id, template_arg)
@@ -108,3 +140,12 @@ async def voice_assistant_stop_to_code(config, action_id, template_arg, args):
     var = cg.new_Pvariable(action_id, template_arg)
     await cg.register_parented(var, config[CONF_ID])
     return var
+
+
+@register_condition(
+    "voice_assistant.is_running", IsRunningCondition, VOICE_ASSISTANT_ACTION_SCHEMA
+)
+async def voice_assistant_is_running_to_code(config, condition_id, template_arg, args):
+    var = cg.new_Pvariable(condition_id, template_arg)
+    await cg.register_parented(var, config[CONF_ID])
+    return var
diff --git a/esphome/components/voice_assistant/voice_assistant.cpp b/esphome/components/voice_assistant/voice_assistant.cpp
index 42455787111f..44d640ff394a 100644
--- a/esphome/components/voice_assistant/voice_assistant.cpp
+++ b/esphome/components/voice_assistant/voice_assistant.cpp
@@ -69,17 +69,42 @@ void VoiceAssistant::setup() {
 
 void VoiceAssistant::loop() {
 #ifdef USE_SPEAKER
-  if (this->speaker_ == nullptr) {
+  if (this->speaker_ != nullptr) {
+    uint8_t buf[1024];
+    auto len = this->socket_->read(buf, sizeof(buf));
+    if (len == -1) {
+      return;
+    }
+    this->speaker_->play(buf, len);
+    this->set_timeout("data-incoming", 200, [this]() {
+      if (this->continuous_) {
+        this->request_start(true);
+      }
+    });
     return;
   }
-
-  uint8_t buf[1024];
-  auto len = this->socket_->read(buf, sizeof(buf));
-  if (len == -1) {
+#endif
+#ifdef USE_MEDIA_PLAYER
+  if (this->media_player_ != nullptr) {
+    if (!this->playing_tts_ ||
+        this->media_player_->state == media_player::MediaPlayerState::MEDIA_PLAYER_STATE_PLAYING) {
+      return;
+    }
+    this->set_timeout("playing-media", 1000, [this]() {
+      this->playing_tts_ = false;
+      if (this->continuous_) {
+        this->request_start(true);
+      }
+    });
     return;
   }
-  this->speaker_->play(buf, len);
 #endif
+  // Set a 1 second timeout to start the voice assistant again.
+  this->set_timeout("continuous-no-sound", 1000, [this]() {
+    if (this->continuous_) {
+      this->request_start(true);
+    }
+  });
 }
 
 void VoiceAssistant::start(struct sockaddr_storage *addr, uint16_t port) {
@@ -100,14 +125,19 @@ void VoiceAssistant::start(struct sockaddr_storage *addr, uint16_t port) {
   }
   this->running_ = true;
   this->mic_->start();
+  this->listening_trigger_->trigger();
 }
 
-void VoiceAssistant::request_start() {
+void VoiceAssistant::request_start(bool continuous) {
   ESP_LOGD(TAG, "Requesting start...");
-  if (!api::global_api_server->start_voice_assistant()) {
+  if (!api::global_api_server->start_voice_assistant(this->conversation_id_)) {
     ESP_LOGW(TAG, "Could not request start.");
     this->error_trigger_->trigger("not-connected", "Could not request start.");
+    this->continuous_ = false;
+    return;
   }
+  this->continuous_ = continuous;
+  this->set_timeout("reset-conversation_id", 5 * 60 * 1000, [this]() { this->conversation_id_ = ""; });
 }
 
 void VoiceAssistant::signal_stop() {
@@ -136,9 +166,18 @@ void VoiceAssistant::on_event(const api::VoiceAssistantEventResponse &msg) {
         return;
       }
       ESP_LOGD(TAG, "Speech recognised as: \"%s\"", text.c_str());
+      this->signal_stop();
       this->stt_end_trigger_->trigger(text);
       break;
     }
+    case api::enums::VOICE_ASSISTANT_INTENT_END: {
+      for (auto arg : msg.data) {
+        if (arg.name == "conversation_id") {
+          this->conversation_id_ = std::move(arg.value);
+        }
+      }
+      break;
+    }
     case api::enums::VOICE_ASSISTANT_TTS_START: {
       std::string text;
       for (auto arg : msg.data) {
@@ -166,6 +205,12 @@ void VoiceAssistant::on_event(const api::VoiceAssistantEventResponse &msg) {
         return;
       }
       ESP_LOGD(TAG, "Response URL: \"%s\"", url.c_str());
+#ifdef USE_MEDIA_PLAYER
+      if (this->media_player_ != nullptr) {
+        this->playing_tts_ = true;
+        this->media_player_->make_call().set_media_url(url).perform();
+      }
+#endif
       this->tts_end_trigger_->trigger(url);
       break;
     }
@@ -184,6 +229,8 @@ void VoiceAssistant::on_event(const api::VoiceAssistantEventResponse &msg) {
         }
       }
       ESP_LOGE(TAG, "Error: %s - %s", code.c_str(), message.c_str());
+      this->continuous_ = false;
+      this->signal_stop();
       this->error_trigger_->trigger(code, message);
     }
     default:
diff --git a/esphome/components/voice_assistant/voice_assistant.h b/esphome/components/voice_assistant/voice_assistant.h
index c1a6e8883b86..b103584509d3 100644
--- a/esphome/components/voice_assistant/voice_assistant.h
+++ b/esphome/components/voice_assistant/voice_assistant.h
@@ -15,6 +15,9 @@
 #ifdef USE_SPEAKER
 #include "esphome/components/speaker/speaker.h"
 #endif
+#ifdef USE_MEDIA_PLAYER
+#include "esphome/components/media_player/media_player.h"
+#endif
 #include "esphome/components/socket/socket.h"
 
 namespace esphome {
@@ -22,8 +25,10 @@ namespace voice_assistant {
 
 // Version 1: Initial version
 // Version 2: Adds raw speaker support
+// Version 3: Adds continuous support
 static const uint32_t INITIAL_VERSION = 1;
 static const uint32_t SPEAKER_SUPPORT = 2;
+static const uint32_t SILENCE_DETECTION_SUPPORT = 3;
 
 class VoiceAssistant : public Component {
  public:
@@ -36,20 +41,34 @@ class VoiceAssistant : public Component {
 #ifdef USE_SPEAKER
   void set_speaker(speaker::Speaker *speaker) { this->speaker_ = speaker; }
 #endif
+#ifdef USE_MEDIA_PLAYER
+  void set_media_player(media_player::MediaPlayer *media_player) { this->media_player_ = media_player; }
+#endif
 
   uint32_t get_version() const {
 #ifdef USE_SPEAKER
-    if (this->speaker_ != nullptr)
+    if (this->speaker_ != nullptr) {
+      if (this->silence_detection_) {
+        return SILENCE_DETECTION_SUPPORT;
+      }
       return SPEAKER_SUPPORT;
+    }
 #endif
     return INITIAL_VERSION;
   }
 
-  void request_start();
+  void request_start(bool continuous = false);
   void signal_stop();
 
   void on_event(const api::VoiceAssistantEventResponse &msg);
 
+  bool is_running() const { return this->running_; }
+  void set_continuous(bool continuous) { this->continuous_ = continuous; }
+  bool is_continuous() const { return this->continuous_; }
+
+  void set_silence_detection(bool silence_detection) { this->silence_detection_ = silence_detection; }
+
+  Trigger<> *get_listening_trigger() const { return this->listening_trigger_; }
   Trigger<> *get_start_trigger() const { return this->start_trigger_; }
   Trigger<std::string> *get_stt_end_trigger() const { return this->stt_end_trigger_; }
   Trigger<std::string> *get_tts_start_trigger() const { return this->tts_start_trigger_; }
@@ -61,6 +80,7 @@ class VoiceAssistant : public Component {
   std::unique_ptr<socket::Socket> socket_ = nullptr;
   struct sockaddr_storage dest_addr_;
 
+  Trigger<> *listening_trigger_ = new Trigger<>();
   Trigger<> *start_trigger_ = new Trigger<>();
   Trigger<std::string> *stt_end_trigger_ = new Trigger<std::string>();
   Trigger<std::string> *tts_start_trigger_ = new Trigger<std::string>();
@@ -72,8 +92,16 @@ class VoiceAssistant : public Component {
 #ifdef USE_SPEAKER
   speaker::Speaker *speaker_{nullptr};
 #endif
+#ifdef USE_MEDIA_PLAYER
+  media_player::MediaPlayer *media_player_{nullptr};
+  bool playing_tts_{false};
+#endif
+
+  std::string conversation_id_{""};
 
   bool running_{false};
+  bool continuous_{false};
+  bool silence_detection_;
 };
 
 template<typename... Ts> class StartAction : public Action<Ts...>, public Parented<VoiceAssistant> {
@@ -81,9 +109,22 @@ template<typename... Ts> class StartAction : public Action<Ts...>, public Parent
   void play(Ts... x) override { this->parent_->request_start(); }
 };
 
+template<typename... Ts> class StartContinuousAction : public Action<Ts...>, public Parented<VoiceAssistant> {
+ public:
+  void play(Ts... x) override { this->parent_->request_start(true); }
+};
+
 template<typename... Ts> class StopAction : public Action<Ts...>, public Parented<VoiceAssistant> {
  public:
-  void play(Ts... x) override { this->parent_->signal_stop(); }
+  void play(Ts... x) override {
+    this->parent_->set_continuous(false);
+    this->parent_->signal_stop();
+  }
+};
+
+template<typename... Ts> class IsRunningCondition : public Condition<Ts...>, public Parented<VoiceAssistant> {
+ public:
+  bool check(Ts... x) override { return this->parent_->is_running() || this->parent_->is_continuous(); }
 };
 
 extern VoiceAssistant *global_voice_assistant;  // NOLINT(cppcoreguidelines-avoid-non-const-global-variables)
diff --git a/esphome/components/wifi/__init__.py b/esphome/components/wifi/__init__.py
index c9da07795cd9..068d0157325f 100644
--- a/esphome/components/wifi/__init__.py
+++ b/esphome/components/wifi/__init__.py
@@ -53,6 +53,9 @@
     "HIGH": WiFiPowerSaveMode.WIFI_POWER_SAVE_HIGH,
 }
 WiFiConnectedCondition = wifi_ns.class_("WiFiConnectedCondition", Condition)
+WiFiEnabledCondition = wifi_ns.class_("WiFiEnabledCondition", Condition)
+WiFiEnableAction = wifi_ns.class_("WiFiEnableAction", automation.Action)
+WiFiDisableAction = wifi_ns.class_("WiFiDisableAction", automation.Action)
 
 
 def validate_password(value):
@@ -253,6 +256,7 @@ def _validate(config):
 
 CONF_OUTPUT_POWER = "output_power"
 CONF_PASSIVE_SCAN = "passive_scan"
+CONF_ENABLE_ON_BOOT = "enable_on_boot"
 CONFIG_SCHEMA = cv.All(
     cv.Schema(
         {
@@ -286,6 +290,7 @@ def _validate(config):
                 "This option has been removed. Please use the [disabled] option under the "
                 "new mdns component instead."
             ),
+            cv.Optional(CONF_ENABLE_ON_BOOT, default=True): cv.boolean,
         }
     ),
     _validate,
@@ -385,6 +390,8 @@ def add_sta(ap, network):
     if CONF_OUTPUT_POWER in config:
         cg.add(var.set_output_power(config[CONF_OUTPUT_POWER]))
 
+    cg.add(var.set_enable_on_boot(config[CONF_ENABLE_ON_BOOT]))
+
     if CORE.is_esp8266:
         cg.add_library("ESP8266WiFi", None)
     elif CORE.is_esp32 and CORE.using_arduino:
@@ -410,3 +417,18 @@ def add_sta(ap, network):
 @automation.register_condition("wifi.connected", WiFiConnectedCondition, cv.Schema({}))
 async def wifi_connected_to_code(config, condition_id, template_arg, args):
     return cg.new_Pvariable(condition_id, template_arg)
+
+
+@automation.register_condition("wifi.enabled", WiFiEnabledCondition, cv.Schema({}))
+async def wifi_enabled_to_code(config, condition_id, template_arg, args):
+    return cg.new_Pvariable(condition_id, template_arg)
+
+
+@automation.register_action("wifi.enable", WiFiEnableAction, cv.Schema({}))
+async def wifi_enable_to_code(config, action_id, template_arg, args):
+    return cg.new_Pvariable(action_id, template_arg)
+
+
+@automation.register_action("wifi.disable", WiFiDisableAction, cv.Schema({}))
+async def wifi_disable_to_code(config, action_id, template_arg, args):
+    return cg.new_Pvariable(action_id, template_arg)
diff --git a/esphome/components/wifi/wifi_component.cpp b/esphome/components/wifi/wifi_component.cpp
index db19f9bcc032..ff621291f067 100644
--- a/esphome/components/wifi/wifi_component.cpp
+++ b/esphome/components/wifi/wifi_component.cpp
@@ -36,9 +36,18 @@ float WiFiComponent::get_setup_priority() const { return setup_priority::WIFI; }
 
 void WiFiComponent::setup() {
   ESP_LOGCONFIG(TAG, "Setting up WiFi...");
+  this->wifi_pre_setup_();
+  if (this->enable_on_boot_) {
+    this->start();
+  } else {
+    this->state_ = WIFI_COMPONENT_STATE_DISABLED;
+  }
+}
+
+void WiFiComponent::start() {
+  ESP_LOGCONFIG(TAG, "Starting WiFi...");
   ESP_LOGCONFIG(TAG, "  Local MAC: %s", get_mac_address_pretty().c_str());
   this->last_connected_ = millis();
-  this->wifi_pre_setup_();
 
   uint32_t hash = this->has_sta() ? fnv1_hash(App.get_compilation_time()) : 88491487UL;
 
@@ -135,6 +144,8 @@ void WiFiComponent::loop() {
       case WIFI_COMPONENT_STATE_OFF:
       case WIFI_COMPONENT_STATE_AP:
         break;
+      case WIFI_COMPONENT_STATE_DISABLED:
+        return;
     }
 
     if (this->has_ap() && !this->ap_setup_) {
@@ -387,6 +398,28 @@ void WiFiComponent::print_connect_params_() {
 #endif
 }
 
+void WiFiComponent::enable() {
+  if (this->state_ != WIFI_COMPONENT_STATE_DISABLED)
+    return;
+
+  ESP_LOGD(TAG, "Enabling WIFI...");
+  this->error_from_callback_ = false;
+  this->state_ = WIFI_COMPONENT_STATE_OFF;
+  this->start();
+}
+
+void WiFiComponent::disable() {
+  if (this->state_ == WIFI_COMPONENT_STATE_DISABLED)
+    return;
+
+  ESP_LOGD(TAG, "Disabling WIFI...");
+  this->state_ = WIFI_COMPONENT_STATE_DISABLED;
+  this->wifi_disconnect_();
+  this->wifi_mode_(false, false);
+}
+
+bool WiFiComponent::is_disabled() { return this->state_ == WIFI_COMPONENT_STATE_DISABLED; }
+
 void WiFiComponent::start_scanning() {
   this->action_started_ = millis();
   ESP_LOGD(TAG, "Starting scan...");
@@ -608,7 +641,7 @@ void WiFiComponent::retry_connect() {
 }
 
 bool WiFiComponent::can_proceed() {
-  if (!this->has_sta()) {
+  if (!this->has_sta() || this->state_ == WIFI_COMPONENT_STATE_DISABLED) {
     return true;
   }
   return this->is_connected();
diff --git a/esphome/components/wifi/wifi_component.h b/esphome/components/wifi/wifi_component.h
index d42be43b2d36..d39b0629902b 100644
--- a/esphome/components/wifi/wifi_component.h
+++ b/esphome/components/wifi/wifi_component.h
@@ -47,6 +47,8 @@ struct SavedWifiSettings {
 enum WiFiComponentState {
   /** Nothing has been initialized yet. Internal AP, if configured, is disabled at this point. */
   WIFI_COMPONENT_STATE_OFF = 0,
+  /** WiFi is disabled. */
+  WIFI_COMPONENT_STATE_DISABLED,
   /** WiFi is in cooldown mode because something went wrong, scanning will begin after a short period of time. */
   WIFI_COMPONENT_STATE_COOLDOWN,
   /** WiFi is in STA-only mode and currently scanning for APs. */
@@ -198,6 +200,9 @@ class WiFiComponent : public Component {
   void set_ap(const WiFiAP &ap);
   WiFiAP get_ap() { return this->ap_; }
 
+  void enable();
+  void disable();
+  bool is_disabled();
   void start_scanning();
   void check_scanning_finished();
   void start_connecting(const WiFiAP &ap, bool two);
@@ -224,6 +229,7 @@ class WiFiComponent : public Component {
   // (In most use cases you won't need these)
   /// Setup WiFi interface.
   void setup() override;
+  void start();
   void dump_config() override;
   /// WIFI setup_priority.
   float get_setup_priority() const override;
@@ -282,6 +288,8 @@ class WiFiComponent : public Component {
 
   int8_t wifi_rssi();
 
+  void set_enable_on_boot(bool enable_on_boot) { this->enable_on_boot_ = enable_on_boot; }
+
  protected:
   static std::string format_mac_addr(const uint8_t mac[6]);
   void setup_ap_config_();
@@ -359,18 +367,30 @@ class WiFiComponent : public Component {
   bool btm_{false};
   bool rrm_{false};
 #endif
+  bool enable_on_boot_;
 };
 
 extern WiFiComponent *global_wifi_component;  // NOLINT(cppcoreguidelines-avoid-non-const-global-variables)
 
 template<typename... Ts> class WiFiConnectedCondition : public Condition<Ts...> {
  public:
-  bool check(Ts... x) override;
+  bool check(Ts... x) override { return global_wifi_component->is_connected(); }
 };
 
-template<typename... Ts> bool WiFiConnectedCondition<Ts...>::check(Ts... x) {
-  return global_wifi_component->is_connected();
-}
+template<typename... Ts> class WiFiEnabledCondition : public Condition<Ts...> {
+ public:
+  bool check(Ts... x) override { return !global_wifi_component->is_disabled(); }
+};
+
+template<typename... Ts> class WiFiEnableAction : public Action<Ts...> {
+ public:
+  void play(Ts... x) override { global_wifi_component->enable(); }
+};
+
+template<typename... Ts> class WiFiDisableAction : public Action<Ts...> {
+ public:
+  void play(Ts... x) override { global_wifi_component->disable(); }
+};
 
 }  // namespace wifi
 }  // namespace esphome
diff --git a/esphome/const.py b/esphome/const.py
index cbc8f428f5dd..470f8a46e5f8 100644
--- a/esphome/const.py
+++ b/esphome/const.py
@@ -399,6 +399,7 @@
 CONF_MDNS = "mdns"
 CONF_MEASUREMENT_DURATION = "measurement_duration"
 CONF_MEASUREMENT_SEQUENCE_NUMBER = "measurement_sequence_number"
+CONF_MEDIA_PLAYER = "media_player"
 CONF_MEDIUM = "medium"
 CONF_MEMORY_BLOCKS = "memory_blocks"
 CONF_METHOD = "method"
diff --git a/requirements.txt b/requirements.txt
index 2b5b2c7e1fc6..b994de1932f9 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -7,11 +7,11 @@ tzlocal==5.0.1    # from time
 tzdata>=2021.1  # from time
 pyserial==3.5
 platformio==6.1.7  # When updating platformio, also update Dockerfile
-esptool==4.5.1
+esptool==4.6
 click==8.1.3
 esphome-dashboard==20230516.0
-aioesphomeapi==13.7.5
-zeroconf==0.62.0
+aioesphomeapi==14.0.0
+zeroconf==0.63.0
 
 # esp-idf requires this, but doesn't bundle it by default
 # https://github.com/espressif/esp-idf/blob/220590d599e134d7a5e7f1e683cc4550349ffbf8/requirements.txt#L24
diff --git a/requirements_optional.txt b/requirements_optional.txt
index 2c734301099d..df6b3b387e7a 100644
--- a/requirements_optional.txt
+++ b/requirements_optional.txt
@@ -1,2 +1,3 @@
 pillow>4.0.0
+cairosvg>=2.2.0
 cryptography>=2.0.0,<4
diff --git a/requirements_test.txt b/requirements_test.txt
index 42aba7ddcf75..c099250bd71f 100644
--- a/requirements_test.txt
+++ b/requirements_test.txt
@@ -6,7 +6,7 @@ pre-commit
 
 # Unit tests
 pytest==7.3.1
-pytest-cov==4.0.0
+pytest-cov==4.1.0
 pytest-mock==3.10.0
 pytest-asyncio==0.21.0
 asyncmock==0.4.2
diff --git a/tests/test2.yaml b/tests/test2.yaml
index 8f5633684837..0dae0b257066 100644
--- a/tests/test2.yaml
+++ b/tests/test2.yaml
@@ -680,6 +680,13 @@ image:
     type: RGB565
     use_transparency: no
 
+  - id: mdi_alert
+    file: mdi:alert-circle-outline
+    resize: 50x50
+  - id: another_alert_icon
+    file: mdi:alert-outline
+    type: BINARY
+
 cap1188:
   id: cap1188_component
   address: 0x29
diff --git a/tests/test6.yaml b/tests/test6.yaml
index 6d956aa9c8b5..6224563a77e7 100644
--- a/tests/test6.yaml
+++ b/tests/test6.yaml
@@ -37,24 +37,24 @@ switch:
     output: pin_4
     id: pin_4_switch
 
-light:
-  - platform: rp2040_pio_led_strip
-    id: led_strip
-    pin: GPIO13
-    num_leds: 60
-    pio: 0
-    rgb_order: GRB
-    chipset: WS2812
-  - platform: rp2040_pio_led_strip
-    id: led_strip_custom_timings
-    pin: GPIO13
-    num_leds: 60
-    pio: 1
-    rgb_order: GRB
-    bit0_high: .1us
-    bit0_low: 1.2us
-    bit1_high: .69us
-    bit1_low: .4us
+#light:
+#  - platform: rp2040_pio_led_strip
+#    id: led_strip
+#    pin: GPIO13
+#    num_leds: 60
+#    pio: 0
+#    rgb_order: GRB
+#    chipset: WS2812
+#  - platform: rp2040_pio_led_strip
+#    id: led_strip_custom_timings
+#    pin: GPIO13
+#    num_leds: 60
+#    pio: 1
+#    rgb_order: GRB
+#    bit0_high: .1us
+#    bit0_low: 1.2us
+#    bit1_high: .69us
+#    bit1_low: .4us
 
 
 sensor: