diff --git a/README.md b/README.md
index f400765f..62aaa114 100644
--- a/README.md
+++ b/README.md
@@ -292,6 +292,18 @@ foreach($stream as $response){
 
 ### `Audio` Resource
 
+#### `speech`
+
+Generates audio from the input text.
+
+```php
+$client->audio()->speech([
+    'model' => 'tts-1',
+    'input' => 'The quick brown fox jumped over the lazy dog.',
+    'voice' => 'alloy',
+]); // audio file content as string
+```
+
 #### `transcribe`
 
 Transcribes audio into the input language.
diff --git a/src/Contracts/Resources/AudioContract.php b/src/Contracts/Resources/AudioContract.php
index f462a8a2..6cdf6b6a 100644
--- a/src/Contracts/Resources/AudioContract.php
+++ b/src/Contracts/Resources/AudioContract.php
@@ -7,6 +7,15 @@
 
 interface AudioContract
 {
+    /**
+     * Generates audio from the input text.
+     *
+     * @see https://platform.openai.com/docs/api-reference/audio/createSpeech
+     *
+     * @param  array<string, mixed>  $parameters
+     */
+    public function speech(array $parameters): string;
+
     /**
      * Transcribes audio into the input language.
      *
diff --git a/src/Resources/Audio.php b/src/Resources/Audio.php
index 391220ba..73d5fd15 100644
--- a/src/Resources/Audio.php
+++ b/src/Resources/Audio.php
@@ -14,6 +14,20 @@ final class Audio implements AudioContract
 {
     use Concerns\Transportable;
 
+    /**
+     * Generates audio from the input text.
+     *
+     * @see https://platform.openai.com/docs/api-reference/audio/createSpeech
+     *
+     * @param  array<string, mixed>  $parameters
+     */
+    public function speech(array $parameters): string
+    {
+        $payload = Payload::create('audio/speech', $parameters);
+
+        return $this->transporter->requestContent($payload);
+    }
+
     /**
      * Transcribes audio into the input language.
      *
diff --git a/src/Testing/Resources/AudioTestResource.php b/src/Testing/Resources/AudioTestResource.php
index 16f4d0ea..29a91fea 100644
--- a/src/Testing/Resources/AudioTestResource.php
+++ b/src/Testing/Resources/AudioTestResource.php
@@ -17,6 +17,11 @@ protected function resource(): string
         return Audio::class;
     }
 
+    public function speech(array $parameters): string
+    {
+        return $this->record(__FUNCTION__, $parameters);
+    }
+
     public function transcribe(array $parameters): TranscriptionResponse
     {
         return $this->record(__FUNCTION__, $parameters);
diff --git a/tests/Fixtures/Audio.php b/tests/Fixtures/Audio.php
index b0800ce3..607bcc0c 100644
--- a/tests/Fixtures/Audio.php
+++ b/tests/Fixtures/Audio.php
@@ -153,3 +153,8 @@ function audioFileResource()
 {
     return fopen(__DIR__.'/audio.mp3', 'r');
 }
+
+function audioFileContent(): string
+{
+    return file_get_contents(__DIR__.'/audio.mp3');
+}
diff --git a/tests/Resources/Audio.php b/tests/Resources/Audio.php
index 3a57b1a1..dd482c19 100644
--- a/tests/Resources/Audio.php
+++ b/tests/Resources/Audio.php
@@ -191,3 +191,21 @@
     expect($result->meta())
         ->toBeInstanceOf(MetaInformation::class);
 });
+
+test('text to speech', function () {
+    $client = mockContentClient('POST', 'audio/speech', [
+        'model' => 'tts-1',
+        'input' => 'Hello, how are you?',
+        'voice' => 'alloy',
+    ], audioFileContent());
+
+    $result = $client->audio()->speech([
+        'model' => 'tts-1',
+        'input' => 'Hello, how are you?',
+        'voice' => 'alloy',
+    ]);
+
+    expect($result)
+        ->toBeString()
+        ->toBe(audioFileContent());
+});
diff --git a/tests/Testing/Resources/AudioTestResource.php b/tests/Testing/Resources/AudioTestResource.php
index 01591cb8..f1f57532 100644
--- a/tests/Testing/Resources/AudioTestResource.php
+++ b/tests/Testing/Resources/AudioTestResource.php
@@ -5,6 +5,27 @@
 use OpenAI\Responses\Audio\TranslationResponse;
 use OpenAI\Testing\ClientFake;
 
+it('records a speech request', function () {
+    $fake = new ClientFake([
+        'fake-mp3-content',
+    ]);
+
+    $fake->audio()->speech([
+        'model' => 'tts-1',
+        'input' => 'Hello, how are you?',
+        'voice' => 'alloy',
+    ]);
+
+    $fake->assertSent(Audio::class, function ($method, $parameters) {
+        return $method === 'speech' &&
+            $parameters === [
+                'model' => 'tts-1',
+                'input' => 'Hello, how are you?',
+                'voice' => 'alloy',
+            ];
+    });
+});
+
 it('records an audio transcription request', function () {
     $fake = new ClientFake([
         TranscriptionResponse::fake(),