text-to-speech

Transforming Text to Speech Effortlessly with Llasa

27 Apr 2025
Transforming Text to Speech Effortlessly with Llasa

In a world where communication is becoming increasingly digital, the ability to convert text into natural-sounding speech is invaluable. Llasa provides a robust solution for developers looking to integrate text-to-speech capabilities into their applications. By extending the LLaMA model with advanced speech tokens from the XCodec2 codebook, Llasa can generate high-quality speech from text or speech prompts. With training on 250,000 hours of Chinese-English speech data, this service ensures clarity, speed, and versatility in audio generation.

Imagine the possibilities: from creating engaging voiceovers for educational content to enabling accessibility features in apps, Llasa's text-to-speech capabilities can enhance user experiences significantly. Whether you're building a language learning app, an audiobook platform, or a virtual assistant, integrating Llasa can streamline your development process and provide your users with lifelike auditory experiences.

Prerequisites

To get started with Llasa, you'll need a Cognitive Actions API key and a basic understanding of how to make API calls.

Convert Text to Speech with Llasa

This action allows developers to perform text-to-speech conversion seamlessly. It solves the problem of generating high-quality, human-like speech from written text, making it ideal for a variety of applications.

Input Requirements

The input for this action requires a composite request that includes:

  • text: The text you wish to convert into speech. For example, "为所有的猫猫奋斗终身！" (Translation: "Fight for all cats for a lifetime!"). Ensure the text is clear to facilitate accurate conversion.
  • voiceSample: A URI pointing to a 16kHz voice sample audio file that will be used to create the speech representation. This is essential for the action to function properly. Example: "https://replicate.delivery/pbxt/MiFpnTHt7iIQ8LELP7yEKUvk1yO3HZwz9NquUVpOQ7SNPa74/zero_shot_prompt.wav".
  • promptText (optional): An additional text prompt that can help in processing if provided.

Expected Output

The output will be a URI link to the generated speech audio file, allowing you to easily access and play the audio. For instance, a sample output might look like: "https://assets.cognitiveactions.com/invocations/04b1e0a0-eb82-457a-a306-2e4d1012efd7/7eaac2b1-29db-4449-8a03-a96aa48ce2ce.wav".

Use Cases for this specific action

  • Educational Apps: Enhance learning experiences by providing audio narration for text-based content, making it more engaging for users.
  • Accessibility Features: Implement voice capabilities in applications to assist users with visual impairments, ensuring that content is accessible to everyone.
  • Content Creation: Generate voiceovers for videos, podcasts, or audiobooks, allowing creators to produce high-quality audio content without extensive studio time.
import requests
import json

# Replace with your actual Cognitive Actions API key and endpoint
# Ensure your environment securely handles the API key
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
# This endpoint URL is hypothetical and should be documented for users
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"

action_id = "acd63f85-303e-4381-9295-c4d39bd4a065" # Action ID for: Convert Text to Speech with Llasa

# Construct the exact input payload based on the action's requirements
# This example uses the predefined example_input for this action:
payload = {
  "text": "为所有的猫猫奋斗终身！",
  "voiceSample": "https://replicate.delivery/pbxt/MiFpnTHt7iIQ8LELP7yEKUvk1yO3HZwz9NquUVpOQ7SNPa74/zero_shot_prompt.wav"
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json",
    # Add any other required headers for the Cognitive Actions API
}

# Prepare the request body for the hypothetical execution endpoint
request_body = {
    "action_id": action_id,
    "inputs": payload
}

print(f"--- Calling Cognitive Action: {action.name or action_id} ---")
print(f"Endpoint: {COGNITIVE_ACTIONS_EXECUTE_URL}")
print(f"Action ID: {action_id}")
print("Payload being sent:")
print(json.dumps(request_body, indent=2))
print("------------------------------------------------")

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json=request_body
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully. Result:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body (non-JSON): {e.response.text}")
    print("------------------------------------------------")

Conclusion

Llasa's text-to-speech capabilities open up a new world of opportunities for developers. By providing a straightforward method to convert text into lifelike speech, it not only enhances user engagement but also improves accessibility. Whether you are looking to create educational tools, enhance user interaction, or streamline content production, integrating Llasa can significantly boost your application's value. Start exploring the power of text-to-speech today and see how it can transform your projects!

← Back to All Articles