diff --git a/package-lock.json b/package-lock.json
index fedd0d066c6..21ad7b77725 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -24,6 +24,8 @@
 				"date-fns": "^2.29.3",
 				"dotenv": "^16.0.3",
 				"express": "^4.21.2",
+				"extendable-media-recorder": "^9.2.19",
+				"extendable-media-recorder-wav-encoder": "^7.0.123",
 				"file-type": "^19.4.1",
 				"google-auth-library": "^9.13.0",
 				"handlebars": "^4.7.8",
@@ -55,6 +57,7 @@
 				"tailwind-scrollbar": "^3.0.0",
 				"tailwindcss": "^3.4.0",
 				"uuid": "^10.0.0",
+				"wavefile": "^11.0.0",
 				"zod": "^3.22.3"
 			},
 			"devDependencies": {
@@ -1320,9 +1323,10 @@
 			}
 		},
 		"node_modules/@babel/runtime": {
-			"version": "7.25.6",
-			"resolved": "https://registry.npmjs.org/@babel/runtime/-/runtime-7.25.6.tgz",
-			"integrity": "sha512-VBj9MYyDb9tuLq7yzqjgzt6Q+IBQLrGZfdjOekyEirZPHxXWoTSGUTMrpsfi58Up73d13NfYLv8HT9vmznjzhQ==",
+			"version": "7.26.7",
+			"resolved": "https://registry.npmjs.org/@babel/runtime/-/runtime-7.26.7.tgz",
+			"integrity": "sha512-AOPI3D+a8dXnja+iwsUqGRjr1BbZIe771sXdapOtYI531gSqpi92vXivKcq2asu/DFpdl1ceFAKZyRzK2PCVcQ==",
+			"license": "MIT",
 			"dependencies": {
 				"regenerator-runtime": "^0.14.0"
 			},
@@ -5011,6 +5015,19 @@
 				"node": ">=8.0.0"
 			}
 		},
+		"node_modules/automation-events": {
+			"version": "7.1.5",
+			"resolved": "https://registry.npmjs.org/automation-events/-/automation-events-7.1.5.tgz",
+			"integrity": "sha512-1t/XcUE/nhHtAyePNZHQMU3/Ka9E0CB1aA8FpzSwc1EBHMsZBtX6H5sBNngCL7pmlxoaDrpcwwHE7Kd6lkYoWA==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"tslib": "^2.8.1"
+			},
+			"engines": {
+				"node": ">=18.2.0"
+			}
+		},
 		"node_modules/autoprefixer": {
 			"version": "10.4.20",
 			"resolved": "https://registry.npmjs.org/autoprefixer/-/autoprefixer-10.4.20.tgz",
@@ -5243,6 +5260,18 @@
 				"node": ">=8"
 			}
 		},
+		"node_modules/broker-factory": {
+			"version": "3.0.109",
+			"resolved": "https://registry.npmjs.org/broker-factory/-/broker-factory-3.0.109.tgz",
+			"integrity": "sha512-yl0S910a6rStKy5Pb4VxtpgyrY0vWoSoXRnAou22jUlK0ikakBSeIh1M4wSl3PgZ8BvYQ9Vt2o3L7hIp9738nw==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"fast-unique-numbers": "^9.0.15",
+				"tslib": "^2.8.1",
+				"worker-factory": "^7.0.36"
+			}
+		},
 		"node_modules/brotli": {
 			"version": "1.3.3",
 			"resolved": "https://registry.npmjs.org/brotli/-/brotli-1.3.3.tgz",
@@ -6884,6 +6913,56 @@
 			"resolved": "https://registry.npmjs.org/extend/-/extend-3.0.2.tgz",
 			"integrity": "sha512-fjquC59cD7CyW6urNXK0FBufkZcoiGG80wTuPujX590cB5Ttln20E2UB4S/WARVqhXffZl2LNgS+gQdPIIim/g=="
 		},
+		"node_modules/extendable-media-recorder": {
+			"version": "9.2.19",
+			"resolved": "https://registry.npmjs.org/extendable-media-recorder/-/extendable-media-recorder-9.2.19.tgz",
+			"integrity": "sha512-7/YoyoqGGaphJkP6NgxgrDBAE4EfJF6CODwL0siKvJnQZ6YMYbUMdCS5sNRfIDqcnd5oR7I6EpsJHvLl+7R4AQ==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"media-encoder-host": "^9.0.14",
+				"multi-buffer-data-view": "^6.0.17",
+				"recorder-audio-worklet": "^6.0.41",
+				"standardized-audio-context": "^25.3.77",
+				"subscribable-things": "^2.1.47",
+				"tslib": "^2.8.1"
+			}
+		},
+		"node_modules/extendable-media-recorder-wav-encoder": {
+			"version": "7.0.123",
+			"resolved": "https://registry.npmjs.org/extendable-media-recorder-wav-encoder/-/extendable-media-recorder-wav-encoder-7.0.123.tgz",
+			"integrity": "sha512-tQFU8UD0HiQNI4F7dzPXR4gskXijpn9dv/wJUH0l+CGCZN4Q8WcEYJPI8CYxHqGOeMaFzjvBlqkkSIFZuEeVMg==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"extendable-media-recorder-wav-encoder-broker": "^7.0.113",
+				"extendable-media-recorder-wav-encoder-worker": "^8.0.110",
+				"tslib": "^2.8.1"
+			}
+		},
+		"node_modules/extendable-media-recorder-wav-encoder-broker": {
+			"version": "7.0.113",
+			"resolved": "https://registry.npmjs.org/extendable-media-recorder-wav-encoder-broker/-/extendable-media-recorder-wav-encoder-broker-7.0.113.tgz",
+			"integrity": "sha512-KkgSfoG7UKxK9tXQa4d/AsDf92bY2FTbg9H6GWMbbOA3IRuChTd70hBwPAk/WNsqPfGUxgJ8Swdq0PX43q9+mg==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"broker-factory": "^3.0.109",
+				"extendable-media-recorder-wav-encoder-worker": "^8.0.110",
+				"tslib": "^2.8.1"
+			}
+		},
+		"node_modules/extendable-media-recorder-wav-encoder-worker": {
+			"version": "8.0.110",
+			"resolved": "https://registry.npmjs.org/extendable-media-recorder-wav-encoder-worker/-/extendable-media-recorder-wav-encoder-worker-8.0.110.tgz",
+			"integrity": "sha512-EvpJglMZuqkOu7HErjvV7hn9BtDSuo2PKaw8cw3oWqzXhKn8NdhX5hkqTpSWuXObfM1bnxxsUYeenG5Yd6z2Ew==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"tslib": "^2.8.1",
+				"worker-factory": "^7.0.36"
+			}
+		},
 		"node_modules/fast-copy": {
 			"version": "3.0.2",
 			"resolved": "https://registry.npmjs.org/fast-copy/-/fast-copy-3.0.2.tgz",
@@ -6952,6 +7031,19 @@
 			"resolved": "https://registry.npmjs.org/fast-safe-stringify/-/fast-safe-stringify-2.1.1.tgz",
 			"integrity": "sha512-W+KJc2dmILlPplD/H4K9l9LcAHAfPtP6BY84uVLXQ6Evcz9Lcg33Y2z1IVblT6xdY54PXYVHEv+0Wpq8Io6zkA=="
 		},
+		"node_modules/fast-unique-numbers": {
+			"version": "9.0.15",
+			"resolved": "https://registry.npmjs.org/fast-unique-numbers/-/fast-unique-numbers-9.0.15.tgz",
+			"integrity": "sha512-vHj0sfq6yB37b/RAAsAJ2DzIp0LR5NlUit7nYFp2YfTUcKL9m/Yk0f0kvYPV4oiuFYXdtO5scs3LQX7qiPAVYQ==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"tslib": "^2.8.1"
+			},
+			"engines": {
+				"node": ">=18.2.0"
+			}
+		},
 		"node_modules/fast-xml-parser": {
 			"version": "4.4.1",
 			"resolved": "https://registry.npmjs.org/fast-xml-parser/-/fast-xml-parser-4.4.1.tgz",
@@ -8857,6 +8949,43 @@
 			"integrity": "sha512-GaqWWShW4kv/G9IEucWScBx9G1/vsFZZJUO+tD26M8J8z3Kw5RDQjaoZe03YAClgeS/SWPOcb4nkFBTEi5DUEA==",
 			"dev": true
 		},
+		"node_modules/media-encoder-host": {
+			"version": "9.0.14",
+			"resolved": "https://registry.npmjs.org/media-encoder-host/-/media-encoder-host-9.0.14.tgz",
+			"integrity": "sha512-paFcVqASpwhpOhvopkzNBRtNd7BQtZic8Eq0GggyImTMcN6xXS0ixJ5egEOEVrHHws/fAoRH0j39ddX5EWQznw==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"media-encoder-host-broker": "^8.0.13",
+				"media-encoder-host-worker": "^10.0.13",
+				"tslib": "^2.8.1"
+			}
+		},
+		"node_modules/media-encoder-host-broker": {
+			"version": "8.0.13",
+			"resolved": "https://registry.npmjs.org/media-encoder-host-broker/-/media-encoder-host-broker-8.0.13.tgz",
+			"integrity": "sha512-mcbATyu/NbebqNQYQypcwuDJ3sCZf3EQtlUMm4viZQ4aMpNy85mte/E3WGzA+VZNKdWR7IavmvmhO2lgJWpYQA==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"broker-factory": "^3.0.109",
+				"fast-unique-numbers": "^9.0.15",
+				"media-encoder-host-worker": "^10.0.13",
+				"tslib": "^2.8.1"
+			}
+		},
+		"node_modules/media-encoder-host-worker": {
+			"version": "10.0.13",
+			"resolved": "https://registry.npmjs.org/media-encoder-host-worker/-/media-encoder-host-worker-10.0.13.tgz",
+			"integrity": "sha512-Yg1+xi8203jcpWYj7/o1JTy0k4MBTpZGo9tG4EwEb/Est1iWXhdKhVYjN4Uj/E3CCgrXr8WNO5ih47cWRSSWtw==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"extendable-media-recorder-wav-encoder-broker": "^7.0.113",
+				"tslib": "^2.8.1",
+				"worker-factory": "^7.0.36"
+			}
+		},
 		"node_modules/media-typer": {
 			"version": "0.3.0",
 			"resolved": "https://registry.npmjs.org/media-typer/-/media-typer-0.3.0.tgz",
@@ -9422,6 +9551,19 @@
 				"url": "https://github.com/sponsors/sindresorhus"
 			}
 		},
+		"node_modules/multi-buffer-data-view": {
+			"version": "6.0.17",
+			"resolved": "https://registry.npmjs.org/multi-buffer-data-view/-/multi-buffer-data-view-6.0.17.tgz",
+			"integrity": "sha512-L76daSLXwMS1T8uL6RUmYyVL3L5hGor5UwDoTqFBYwlnvhtfpvVS6lu5oePrXi0B+X2JKTXZxBerqqb4Q7fM+g==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"tslib": "^2.8.1"
+			},
+			"engines": {
+				"node": ">=18.2.0"
+			}
+		},
 		"node_modules/mute-stream": {
 			"version": "2.0.0",
 			"resolved": "https://registry.npmjs.org/mute-stream/-/mute-stream-2.0.0.tgz",
@@ -10837,6 +10979,32 @@
 				"node": ">= 12.13.0"
 			}
 		},
+		"node_modules/recorder-audio-worklet": {
+			"version": "6.0.41",
+			"resolved": "https://registry.npmjs.org/recorder-audio-worklet/-/recorder-audio-worklet-6.0.41.tgz",
+			"integrity": "sha512-sCgsdBgAMlD5dbJ6NRFOyCZsYuS7y7Twg/p8Eff8gy9FOZIxFiYdxu/s/1oLuMiTkSsR5K9ejPSN5vthYGtSFQ==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"broker-factory": "^3.0.109",
+				"fast-unique-numbers": "^9.0.15",
+				"recorder-audio-worklet-processor": "^5.0.30",
+				"standardized-audio-context": "^25.3.77",
+				"subscribable-things": "^2.1.47",
+				"tslib": "^2.8.1",
+				"worker-factory": "^7.0.36"
+			}
+		},
+		"node_modules/recorder-audio-worklet-processor": {
+			"version": "5.0.30",
+			"resolved": "https://registry.npmjs.org/recorder-audio-worklet-processor/-/recorder-audio-worklet-processor-5.0.30.tgz",
+			"integrity": "sha512-HL6ySdEbD6f909MLG5S7qg2gZBzxxHI9F/P5G9dMEDm3mejzZQx9zvP+Uw0SCIBmz+mztOQOkxCF3tc6FLCe3w==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"tslib": "^2.8.1"
+			}
+		},
 		"node_modules/regenerator-runtime": {
 			"version": "0.14.1",
 			"resolved": "https://registry.npmjs.org/regenerator-runtime/-/regenerator-runtime-0.14.1.tgz",
@@ -11044,6 +11212,12 @@
 				"queue-microtask": "^1.2.2"
 			}
 		},
+		"node_modules/rxjs-interop": {
+			"version": "2.0.0",
+			"resolved": "https://registry.npmjs.org/rxjs-interop/-/rxjs-interop-2.0.0.tgz",
+			"integrity": "sha512-ASEq9atUw7lualXB+knvgtvwkCEvGWV2gDD/8qnASzBkzEARZck9JAyxmY8OS6Nc1pCPEgDTKNcx+YqqYfzArw==",
+			"license": "MIT"
+		},
 		"node_modules/sade": {
 			"version": "1.8.1",
 			"resolved": "https://registry.npmjs.org/sade/-/sade-1.8.1.tgz",
@@ -11585,6 +11759,17 @@
 			"integrity": "sha512-1XMJE5fQo1jGH6Y/7ebnwPOBEkIEnT4QF32d5R1+VXdXveM0IBMJt8zfaxX1P3QhVwrYe+576+jkANtSS2mBbw==",
 			"dev": true
 		},
+		"node_modules/standardized-audio-context": {
+			"version": "25.3.77",
+			"resolved": "https://registry.npmjs.org/standardized-audio-context/-/standardized-audio-context-25.3.77.tgz",
+			"integrity": "sha512-Ki9zNz6pKcC5Pi+QPjPyVsD9GwJIJWgryji0XL9cAJXMGyn+dPOf6Qik1AHei0+UNVcc4BOCa0hWLBzlwqsW/A==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.25.6",
+				"automation-events": "^7.0.9",
+				"tslib": "^2.7.0"
+			}
+		},
 		"node_modules/static-eval": {
 			"version": "2.0.2",
 			"resolved": "https://registry.npmjs.org/static-eval/-/static-eval-2.0.2.tgz",
@@ -11800,6 +11985,17 @@
 				"url": "https://github.com/sponsors/Borewit"
 			}
 		},
+		"node_modules/subscribable-things": {
+			"version": "2.1.47",
+			"resolved": "https://registry.npmjs.org/subscribable-things/-/subscribable-things-2.1.47.tgz",
+			"integrity": "sha512-LEBz6py2ZombbDAziKv9mBI5n/CFb4kZ2KGyLwoazB/G91x+7s1st/rE1YAIPj/kjCEPmry9BjnRQ7NCBcq85A==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"rxjs-interop": "^2.0.0",
+				"tslib": "^2.8.1"
+			}
+		},
 		"node_modules/sucrase": {
 			"version": "3.35.0",
 			"resolved": "https://registry.npmjs.org/sucrase/-/sucrase-3.35.0.tgz",
@@ -12468,9 +12664,10 @@
 			"devOptional": true
 		},
 		"node_modules/tslib": {
-			"version": "2.7.0",
-			"resolved": "https://registry.npmjs.org/tslib/-/tslib-2.7.0.tgz",
-			"integrity": "sha512-gLXCKdN1/j47AiHiOkJN69hJmcbGTHI0ImLmbYLHykhgeN0jVGola9yVjFgzCUklsZQMW55o+dW7IXv3RCXDzA=="
+			"version": "2.8.1",
+			"resolved": "https://registry.npmjs.org/tslib/-/tslib-2.8.1.tgz",
+			"integrity": "sha512-oJFu94HQb+KVduSUQL7wnpmqnfmLsOA/nAh6b6EH0wCEoK0/mPeXU6c3wKDV83MkOuHPRHtSXKKU99IBazS/2w==",
+			"license": "0BSD"
 		},
 		"node_modules/type-check": {
 			"version": "0.4.0",
@@ -12971,6 +13168,18 @@
 				"node": ">=14"
 			}
 		},
+		"node_modules/wavefile": {
+			"version": "11.0.0",
+			"resolved": "https://registry.npmjs.org/wavefile/-/wavefile-11.0.0.tgz",
+			"integrity": "sha512-/OBiAALgWU24IG7sC84cDO/KfFuvajWc5Uec0oV2zrpOOZZDgGdOwHwgEzOrwh8jkubBk7PtZfQBIcI1OaE5Ng==",
+			"license": "MIT",
+			"bin": {
+				"wavefile": "bin/wavefile.js"
+			},
+			"engines": {
+				"node": ">=8"
+			}
+		},
 		"node_modules/web-streams-polyfill": {
 			"version": "4.0.0-beta.3",
 			"resolved": "https://registry.npmjs.org/web-streams-polyfill/-/web-streams-polyfill-4.0.0-beta.3.tgz",
@@ -13079,6 +13288,17 @@
 			"resolved": "https://registry.npmjs.org/wordwrap/-/wordwrap-1.0.0.tgz",
 			"integrity": "sha512-gvVzJFlPycKc5dZN4yPkP8w7Dc37BtP1yczEneOb4uq34pXZcvrtRTmWV8W+Ume+XCxKgbjM+nevkyFPMybd4Q=="
 		},
+		"node_modules/worker-factory": {
+			"version": "7.0.36",
+			"resolved": "https://registry.npmjs.org/worker-factory/-/worker-factory-7.0.36.tgz",
+			"integrity": "sha512-gSYPHZRUBXUkON4LSyA7/GAVmBuJDU1WeJhFFp+ZmlxNKInWCvxG1UAS5KnwcayTfGHGVEvM1BnfgmUu8OtpnA==",
+			"license": "MIT",
+			"dependencies": {
+				"@babel/runtime": "^7.26.0",
+				"fast-unique-numbers": "^9.0.15",
+				"tslib": "^2.8.1"
+			}
+		},
 		"node_modules/wrap-ansi": {
 			"version": "8.1.0",
 			"resolved": "https://registry.npmjs.org/wrap-ansi/-/wrap-ansi-8.1.0.tgz",
diff --git a/package.json b/package.json
index 38fe6c3b834..8edf3535325 100644
--- a/package.json
+++ b/package.json
@@ -77,6 +77,8 @@
 		"date-fns": "^2.29.3",
 		"dotenv": "^16.0.3",
 		"express": "^4.21.2",
+		"extendable-media-recorder": "^9.2.19",
+		"extendable-media-recorder-wav-encoder": "^7.0.123",
 		"file-type": "^19.4.1",
 		"google-auth-library": "^9.13.0",
 		"handlebars": "^4.7.8",
@@ -108,6 +110,7 @@
 		"tailwind-scrollbar": "^3.0.0",
 		"tailwindcss": "^3.4.0",
 		"uuid": "^10.0.0",
+		"wavefile": "^11.0.0",
 		"zod": "^3.22.3"
 	},
 	"optionalDependencies": {
diff --git a/src/lib/components/chat/ChatTranscribe.svelte b/src/lib/components/chat/ChatTranscribe.svelte
new file mode 100644
index 00000000000..08d67439d38
--- /dev/null
+++ b/src/lib/components/chat/ChatTranscribe.svelte
@@ -0,0 +1,110 @@
+<script lang="ts">
+	import { error } from "$lib/stores/errors";
+	import CarbonMicrophone from "~icons/carbon/microphone";
+	import EosIconsLoading from "~icons/eos-icons/loading";
+	import TranscribeWorker from "$lib/workers/transcribeWorker?worker";
+
+	onMount(async () => {
+		const { connect } = await import("extendable-media-recorder-wav-encoder");
+		const { register } = await import("extendable-media-recorder");
+
+		await register(await connect());
+	});
+
+	import type { MediaRecorder as ExtendableMediaRecorder } from "extendable-media-recorder";
+	import { onMount } from "svelte";
+
+	export let value = "";
+
+	$: downloading = false;
+	$: processing = false;
+	$: recording = false;
+
+	$: progress = 0;
+
+	const worker = new TranscribeWorker();
+	const userLanguage = navigator.language.slice(0, 2).toLowerCase();
+
+	worker.onmessage = (ev) => {
+		if (ev.data.type === "transcription") {
+			value += ev.data.transcription;
+		} else if (ev.data.type === "progress") {
+			progress = ev.data.progress;
+		} else if (ev.data.type === "error") {
+			error.set(ev.data.error);
+		} else if (ev.data.type === "update") {
+			if (ev.data.message === "downloading") {
+				downloading = true;
+			} else if (ev.data.message === "ready") {
+				downloading = false;
+			}
+		}
+	};
+
+	let mediaRecorder: InstanceType<typeof ExtendableMediaRecorder> | null = null;
+
+	$: lastWordFromValue = value.split(" ").pop();
+
+	async function onClick() {
+		if (mediaRecorder) {
+			recording = true;
+			mediaRecorder.start();
+		} else {
+			const { MediaRecorder } = await import("extendable-media-recorder");
+			const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
+			mediaRecorder = new MediaRecorder(stream, {
+				mimeType: "audio/wav",
+			});
+
+			mediaRecorder.ondataavailable = (ev) => {
+				worker.postMessage({
+					type: "audio",
+					audioUrl: URL.createObjectURL(ev.data),
+					language: userLanguage,
+				});
+			};
+
+			mediaRecorder.onstop = () => {
+				recording = false;
+			};
+
+			recording = true;
+			mediaRecorder.start();
+		}
+	}
+
+	$: buttonClasses = [
+		...(recording
+			? ["text-gray-800", "dark:text-gray-200"]
+			: ["text-gray-400", "dark:text-gray-500"]),
+		...(downloading ? ["opacity-20"] : []),
+	];
+</script>
+
+{#if recording}
+	<div
+		class="bg-gray-800/501 absolute left-0 top-0 z-10 flex h-full w-full items-center justify-center rounded-xl p-2.5"
+	>
+		<span class="font-bold text-white">{lastWordFromValue}</span>
+		<button class="btn" on:click|preventDefault={() => mediaRecorder?.stop()}>Stop</button>
+	</div>
+{/if}
+<div class="absolute right-2 top-2 flex size-7 items-center justify-center">
+	<button class="btn relative" on:click|preventDefault={onClick} disabled={downloading}>
+		{#if downloading}
+			<div class="absolute left-1/2 top-1/2 -translate-x-1/2 -translate-y-1/2">
+				<div class="h-1 w-5 rounded-full bg-gray-200 dark:bg-gray-700">
+					<div
+						class="h-full rounded-full bg-blue-400 transition-all duration-300 dark:bg-blue-500"
+						style="width: {progress * 100}%;"
+					/>
+				</div>
+			</div>
+		{/if}
+		{#if processing}
+			<EosIconsLoading class="size-5" />
+		{:else}
+			<CarbonMicrophone class={buttonClasses.join(" ")} />
+		{/if}
+	</button>
+</div>
diff --git a/src/lib/components/chat/ChatWindow.svelte b/src/lib/components/chat/ChatWindow.svelte
index 761fb8b2dac..9e1c256e2d2 100644
--- a/src/lib/components/chat/ChatWindow.svelte
+++ b/src/lib/components/chat/ChatWindow.svelte
@@ -35,6 +35,8 @@
 	import { cubicInOut } from "svelte/easing";
 	import type { ToolFront } from "$lib/types/Tool";
 	import { loginModalOpen } from "$lib/stores/loginModal";
+	import ChatTranscribe from "./ChatTranscribe.svelte";
+	import { isWebGPUSupported } from "$lib/utils/isWebGPUSupported";
 
 	export let messages: Message[] = [];
 	export let loading = false;
@@ -440,7 +442,11 @@
 								modelIsMultimodal={currentModel.multimodal}
 							/>
 						{/if}
-
+						{#await isWebGPUSupported() then isWebGPUSupported}
+							{#if browser && isWebGPUSupported}
+								<ChatTranscribe bind:value={message} />
+							{/if}
+						{/await}
 						{#if loading}
 							<button
 								disabled
diff --git a/src/lib/utils/isWebGPUSupported.ts b/src/lib/utils/isWebGPUSupported.ts
new file mode 100644
index 00000000000..7d79bb250b8
--- /dev/null
+++ b/src/lib/utils/isWebGPUSupported.ts
@@ -0,0 +1,10 @@
+export async function isWebGPUSupported() {
+	try {
+		// @ts-expect-error gpu is not yet typed
+		const adapter = await navigator.gpu.requestAdapter();
+		const device = await adapter.requestDevice();
+		return device !== null;
+	} catch (e) {
+		return false;
+	}
+}
diff --git a/src/lib/workers/transcribeWorker.ts b/src/lib/workers/transcribeWorker.ts
new file mode 100644
index 00000000000..45166b70e84
--- /dev/null
+++ b/src/lib/workers/transcribeWorker.ts
@@ -0,0 +1,145 @@
+import { AutomaticSpeechRecognitionPipeline, pipeline } from "@huggingface/transformers";
+import { WaveFile } from "wavefile";
+
+type AudioMessage = {
+	audioUrl: URL;
+	type: "audio";
+	language: string;
+};
+
+type TranscriptionMessage = {
+	transcription: string;
+	type: "transcription";
+};
+
+type DownloadProgressMessage = {
+	progress: number;
+	type: "progress";
+};
+
+type ErrorMessage = {
+	error: string;
+	type: "error";
+};
+
+type UpdateMessage = {
+	message: "downloading" | "ready";
+	type: "update";
+};
+
+type Message =
+	| AudioMessage
+	| TranscriptionMessage
+	| DownloadProgressMessage
+	| ErrorMessage
+	| UpdateMessage;
+
+class TranscriberSingleton {
+	private pipeline: AutomaticSpeechRecognitionPipeline | null = null;
+
+	public async getPipeline() {
+		if (!this.pipeline) {
+			const filesToBeDownloaded: Array<{ file: string; total: number; loaded: number }> = [];
+
+			self.postMessage({
+				message: "downloading",
+				type: "update",
+			});
+
+			this.pipeline = await pipeline(
+				"automatic-speech-recognition",
+				"onnx-community/whisper-base",
+				{
+					device: "webgpu",
+					progress_callback: (ev) => {
+						if (!filesToBeDownloaded.find((f) => f.file === ev.file)) {
+							filesToBeDownloaded.push({
+								file: ev.file,
+								total: ev.total ?? 0,
+								loaded: ev.loaded ?? 0,
+							});
+						} else {
+							const file = filesToBeDownloaded.find((f) => f.file === ev.file);
+							if (file) {
+								file.loaded = ev.loaded ?? 0;
+								file.total = ev.total ?? 0;
+							}
+						}
+
+						if (filesToBeDownloaded.some((f) => f.total === 0) && filesToBeDownloaded.length < 2) {
+							return;
+						}
+
+						const totalLoaded = filesToBeDownloaded.reduce((acc, f) => acc + f.loaded, 0);
+						const totalSize = filesToBeDownloaded.reduce((acc, f) => acc + f.total, 0);
+
+						const newProgress = totalLoaded / totalSize;
+						if (!isNaN(newProgress) && isFinite(newProgress)) {
+							self.postMessage({
+								type: "progress",
+								progress: newProgress,
+							});
+						}
+					},
+				}
+			);
+
+			self.postMessage({
+				message: "ready",
+				type: "update",
+			});
+		}
+		return this.pipeline;
+	}
+}
+
+const transcriberSingleton = new TranscriberSingleton();
+
+async function processAudioUrl(audioUrl: URL): Promise<Float64Array> {
+	const arrayBuffer = await fetch(audioUrl).then((x) => x.arrayBuffer());
+	const buffer = new Uint8Array(arrayBuffer);
+	const wav = new WaveFile(buffer);
+	wav.toBitDepth("32f"); // Pipeline expects input as a Float32Array
+	wav.toSampleRate(16000); // Whisper expects audio with a sampling rate of 16000
+	let audioData = wav.getSamples();
+	if (Array.isArray(audioData)) {
+		if (audioData.length > 1) {
+			const SCALING_FACTOR = Math.sqrt(2);
+
+			// Merge channels (into first channel to save memory)
+			for (let i = 0; i < audioData[0].length; ++i) {
+				audioData[0][i] = (SCALING_FACTOR * (audioData[0][i] + audioData[1][i])) / 2;
+			}
+		}
+
+		audioData = audioData[0];
+	}
+
+	return audioData;
+}
+
+self.onmessage = async (ev: MessageEvent<Message>) => {
+	if (ev.data.type === "audio") {
+		const audioUrl = ev.data.audioUrl;
+		const pipeline = await transcriberSingleton.getPipeline();
+
+		const audioData = await processAudioUrl(audioUrl);
+
+		await pipeline(audioData, {
+			language: ev.data.language,
+			task: "transcribe",
+		})
+			.then((output) => {
+				if (Array.isArray(output)) {
+					return output.map((o) => o.text).join(" ");
+				}
+				return output.text;
+			})
+			.then((transcription) => {
+				self.postMessage({
+					type: "transcription",
+					transcription,
+				});
+			});
+	}
+};