## 🔥 Feature: External API for Audiobook Generation

README.md

@@ -38,6 +38,29 @@ mlx_audio.tts.generate --text "Hello, world" --file_prefix hello
 mlx_audio.tts.generate --text "Hello, world" --speed 1.4
 ```
 
+### How to call from python
+
+To generate audio with an LLM use:
+
+```python
+from mlx_audio.tts.generate import generate_audio
+
+# Example: Generate an audiobook chapter as audio
+generate_audio(
+    text="In the beginning, the universe was created...",
+    model="prince-canuma/Kokoro-82M",
+    voice="af_heart",
+    speed=1.2,
+    lang_code="en",
+    file_path="audiobook_chapter1",
+    audio_format="wav",
+    sample_rate=24000,
+    verbose=True  # Set to False to disable print messages
+)
+
+print("Audiobook chapter successfully generated!")
+
+```
 
 ### Web Interface & API Server
 

mlx_audio/tts/generate.py

@@ -117,5 +117,59 @@ def main():
         traceback.print_exc()
 
 
+def generate_audio(text: str,
+                   model: str = "prince-canuma/Kokoro-82M",
+                   voice: str = "af_heart",
+                   speed: float = 1.0,
+                   lang_code: str = "a",
+                   file_path: str = "audio",
+                   audio_format: str = "wav",
+                   sample_rate: int = 24000,
+                   verbose: bool = True) -> None:
+    """
+    Generates audio from text using a specified TTS model.
+
+    Parameters:
+    - text (str): The input text to be converted to speech.
+    - model (str): The TTS model to use (default: "prince-canuma/Kokoro-82M").
+    - voice (str): The voice style to use (default: "af_heart").
+    - speed (float): Playback speed multiplier (default: 1.0).
+    - lang_code (str): The language code (default: "a").
+    - file_path (str): The output file path without extension (default: "audio").
+    - audio_format (str): Output audio format (e.g., "wav", "flac") (default: "wav").
+    - sample_rate (int): Sampling rate in Hz (default: 24000).
+    - verbose (bool): Whether to print status messages (default: True).
+
+    Returns:
+    - None: The function writes the generated audio to a file.
+    """
+    try:
+        # Load the specified TTS model
+        model_instance = load_model(model_path=model)
+
+        # Generate audio using the model
+        results = model_instance.generate(
+            text=text,
+            voice=voice,
+            speed=speed,
+            lang_code=lang_code,
+            verbose=verbose,  # Pass verbose to model if applicable
+        )
+
+        # Concatenate generated audio segments
+        audio_list = [result.audio for result in results]
+        final_audio = mx.concatenate(audio_list, axis=0)
+
+        # Save the audio to the specified file format
+        output_file = f"{file_path}.{audio_format}"
+        sf.write(output_file, final_audio, sample_rate)
+
+        if verbose:
+            print(f"✅ Audio successfully generated and saved as: {output_file}")
+
+    except Exception as e:
+        print(f"❌ Error generating audio: {e}")
+
+
 if __name__ == "__main__":
     main()