# Conversational AI

> Real-time voice AI platform combining streaming STT, LLM orchestration, TTS, interruption handling, turn detection, and AI voice agent capabilities.

This index is the entry point for LLMs. It links to shared product-level docs and to one `llms.txt` per supported platform. Each linked page also has a `.md` twin (append `.md` to the URL) containing the raw markdown.

## Basics

- [Introduction](https://pre.trtc.io/document/64658.md): Tencent AI real-time conversation solution is an SDK that enables real-time audio and video interaction between AI and users by integrating multiple AI large model services. Designed for scenarios including online education, social entertainment, call centers, efficient office work, and medical assistance, it delivers ultra-low latency with conversation delays under 1,000 ms and supports fast deployment through a code-free playground and comprehensive SDKs and APIs.
- [Activate the Service](https://pre.trtc.io/document/69002.md): The Conversational AI Service SDK provides real-time natural voice interaction capabilities powered by large language models. This guide helps users activate the service, utilize free usage time, and renew it by purchasing an RTC-Engine Monthly Package or enabling Pay-as-you-go billing. Key steps include creating an application in the Tencent RTC console, configuring voice agents, and managing service fees comprising audio call, AI service, and speech-to-text charges.
- [Price](https://pre.trtc.io/document/67833.md): The Conversational AI service within Tencent RTC Engine provides real-time voice interaction between humans and large language models. This overview explains the billing components including audio call fees, Conversational AI service fees, and speech-to-text service fees. It targets developers and users integrating or utilizing Conversational AI, detailing free trial usage, monthly package options, and fee calculation methods. Integration guidance is also referenced for implementation.

## Voice Agent

- [Quick Try (No Code)](https://pre.trtc.io/document/68137.md): Tencent RTC Conversational AI quickstart guide provides a no-code platform for developers to rapidly configure and test conversational AI services by integrating third-party Speech To Text (STT), Large Language Model (LLM), and Text To Speech (TTS) providers. This guide covers steps from console login, resource setup, to starting voice conversations, enabling efficient deployment of AI voice agents with multiple supported providers.
- [Quick Try (AI-assisted)](https://pre.trtc.io/agent-skills): Install the Tencent RTC Agent Skills pack (npx -y @tencent-rtc/trtc-agent-skills@latest add) so Claude Code, Cursor, Codex, or CodeBuddy can scaffold a working Conversational AI voice agent — including STT/LLM/TTS provider wiring — directly from a prompt.
- [Android](https://pre.trtc.io/document/69001.md): AIConversationKit Demo is an Android conversational AI SDK that enables developers to quickly experience a high-quality conversational AI project with a complete UI. This guide covers environment preparation, demo download, configuration with Tencent Cloud credentials, running the demo on Android devices, and starting your first AI dialogue, helping developers integrate and test conversational AI features efficiently.
- [Web](https://pre.trtc.io/document/68999.md): The AIConversationKit Demo project quickstart guide provides step-by-step instructions to run a high-quality conversational AI project with a complete UI interface on the Web platform. It guides users through application creation, service activation, parameter configuration, server-side dependency installation, and frontend code execution to quickly experience the conversational AI features.
- [iOS](https://pre.trtc.io/document/69000.md): AIConversationKit is a conversational AI demo project SDK for iOS that enables developers to quickly experience and run a complete conversational AI interface. This guide covers environment preparation, demo download, configuration with Tencent Cloud credentials, and running the demo using Xcode. Users can start their first dialogue with AI within 20 minutes by following step-by-step instructions.
- [Android](https://pre.trtc.io/document/69004.md): AIConversationKit SDK for Android provides a complete conversational AI interface integration guide. This document helps developers prepare the environment, activate the AI service, download and configure the necessary components, perform user sign-in, and start their first AI conversation within 20 minutes. It supports Android 4.4 and above, enabling rapid deployment of conversational AI features on mobile devices.
- [iOS](https://pre.trtc.io/document/69005.md): AIConversationKit for iOS is a conversational AI SDK that enables developers to quickly integrate and launch conversational AI projects. This guide provides step-by-step instructions including environment setup, service activation, component download, project configuration, user sign-in, and starting your first conversational AI project with a complete UI within 20 minutes.
- [Quick Implementation](https://pre.trtc.io/document/68337.md): The RTC Engine SDK quickstart guide provides developers with step-by-step instructions to implement conversational AI solutions across iOS, Android, Web, H5, and Flutter platforms. It covers integrating the SDK, enabling low-latency audio streaming with speech mode for enhanced denoising, initiating AI conversations via backend APIs, and managing chatbot interactions. This guide helps users rapidly deploy flexible AI-powered voice communication features with real-time voice enhancement and context management.
- [Overview](https://pre.trtc.io/document/69592.md): The Speech-to-Text (STT) module is a key component of the Conversational AI SDK powered by Tencent RTC. It captures user voice streams in real time and converts them into accurate text using ultra-low latency audio pipelines and advanced audio processing technologies. Designed for developers integrating conversational AI, it supports Tencent's built-in STT as well as multiple third-party providers like Azure Speech, Deepgram, and Soniox. This module ensures clear transcription even in noisy environments and offers flexible configuration through the STTConfig object.
- [Tencent](https://pre.trtc.io/document/79672.md): Tencent's in-house Speech-to-Text (STT) engine is integrated within the TRTC platform, providing ultra-low latency speech recognition with advanced audio processing capabilities such as AI noise suppression and echo cancellation. This API reference details the STTConfig parameters used to configure language options and voice activity detection settings, enabling developers to customize speech recognition for multiple languages without requiring additional service accounts.
- [Azure](https://pre.trtc.io/document/79673.md): Microsoft Azure Speech Services SDK offers enterprise-grade speech recognition with support for over 100 languages and customizable speech models. This guide explains how to configure Azure as the speech-to-text engine by passing the required JSON parameters in the STTConfig field of the StartAIConversation API, including authentication details like SubscriptionKey and Region. Follow the steps to enable multilingual speech recognition within your Tencent RTC applications.
- [Deepgram](https://pre.trtc.io/document/79674.md): Deepgram is a real-time transcription SDK built on end-to-end deep learning, offering fast and accurate speech-to-text capabilities optimized for English-heavy scenarios. This documentation provides detailed API reference for configuring Deepgram with the StartAIConversation API, including STTConfig and CustomParam fields required for authentication and model selection. Key parameters such as language code, voice activity detection silence time, and Deepgram-specific API key and model options are explained to enable seamless integration.
- [Soniox](https://pre.trtc.io/document/79675.md): Soniox speech recognition API provides a multilingual, low-latency speech-to-text SDK integration for global voice AI applications. This reference details the STTConfig parameters needed to configure Soniox as the speech-to-text engine, including language settings, silence detection, and authentication keys. It supports code-switching scenarios and offers optional customization through language hints and context objects.
- [Overview](https://pre.trtc.io/document/68338.md): Tencent Real-Time Communication (TRTC) offers a conversational AI framework that integrates various leading large language models (LLMs) including OpenAI-compatible engines and no-code platforms like Dify and Coze. Designed for developers and enterprises, it enables real-time, contextually relevant responses with global latency under one second. The platform supports multiple providers with unified or dedicated configurations and automatically attaches HTTP headers for enhanced request tracking and debugging.
- [OpenAI](https://pre.trtc.io/document/79676.md): OpenAI's GPT series models serve as the core large language model (LLM) engine within Tencent RTC's Conversational AI API. This reference details how to configure parameters such as model type, API key, endpoint URL, streaming options, system prompts, timeout, and conversation history for effective integration. It supports advanced models including GPT-5.5, enabling developers to leverage powerful conversational AI capabilities with low latency and multimodal features.
- [Gemini](https://pre.trtc.io/document/79677.md): The Gemini model API reference details how to configure Google's Gemini large language models for conversational AI using Tencent RTC. It covers the OpenAI-compatible protocol parameters required to integrate Gemini 3.1 pro and related models, enabling fast inference with large context windows. This guide is intended for developers integrating Gemini as the LLM engine in their applications, providing parameter descriptions such as API key, model selection, streaming options, and conversation history settings.
- [MiniMax](https://pre.trtc.io/document/79678.md): MiniMax provides AI models optimized for Chinese-language dialogue and creative content generation, supporting the OpenAI-compatible protocol. This API reference details how to configure MiniMax as the large language model engine within Tencent RTC's Conversational AI, including required parameters such as model selection, API keys, and streaming options. It is designed for developers integrating advanced Chinese-language AI capabilities into their applications.
- [Tencent](https://pre.trtc.io/document/79679.md): Tencent Hunyuan is a large language model (LLM) engine integrated into the StartAIConversation API, providing conversational AI capabilities. This API reference details the required and optional parameters for configuring Tencent Hunyuan, including model selection, API key, endpoint URL, streaming options, system prompts, timeout settings, conversation history, and custom metadata. It supports OpenAI-compatible endpoints and enables developers to customize AI interactions effectively.
- [Dify](https://pre.trtc.io/document/79680.md): Dify is an open-source large language model (LLM) application platform that provides an SDK for building AI workflows without coding. It supports complex AI agent logic such as knowledge bases, multi-step workflows, and tool integrations. This document details how to configure Dify as the LLM engine in the StartAIConversation API, including required parameters like APIKey, APIUrl, and options for streaming and session management.
- [Coze](https://pre.trtc.io/document/79681.md): Coze by ByteDance is a no-code AI bot platform SDK that allows developers to create and manage conversational bots using built-in plugins, knowledge bases, and memory through a visual interface. This documentation explains how to integrate Coze as the LLM engine in Tencent RTC's StartAIConversation API by configuring key parameters such as APIKey, BotId, and APIUrl. It provides detailed usage instructions and parameter references to enable rapid prototyping of AI bots with streaming and session management capabilities.
- [Overview](https://pre.trtc.io/document/68340.md): The Text-to-Speech (TTS) module is a real-time speech synthesis SDK by Tencent RTC that converts LLM-generated text into natural-sounding audio with ultra-low latency under 300 ms. It supports TRTC's built-in voice library requiring no external accounts, as well as integration with multiple third-party TTS providers like MiniMax, Azure, Cartesia, ElevenLabs, and Inworld. This flexible framework enables developers to deliver expressive, real-time speech responses across various languages and voice styles.
- [Tencent](https://pre.trtc.io/document/79682.md): Tencent Real-Time Communication (TRTC) offers a native real-time Text-to-Speech (TTS) API with a curated premium voice library supporting Chinese, English (US), and Cantonese languages. This API enables developers to integrate high-quality TTS with minimal overhead and low latency by configuring parameters such as voice selection, speech speed, volume, pitch, and language through the TTSConfig field in the StartAIConversation API. The voice library provides a variety of premium voices for different use cases without requiring external accounts or API keys.
- [MiniMax](https://pre.trtc.io/document/79683.md): MiniMax TTS is a text-to-speech SDK that offers expressive and emotionally rich Chinese voice synthesis with advanced prosody control. It is designed for developers integrating natural-sounding speech with customizable voice parameters such as model type, voice, speed, and authentication keys. This reference details the TTSConfig parameters required to use MiniMax in the StartAIConversation API, including model options, API endpoints, and rate limits for different account types.
- [Azure](https://pre.trtc.io/document/79684.md): Azure Speech Services Text-to-Speech API offers over 400 neural voices in more than 140 languages, enabling multilingual speech synthesis with fine-grained control via SSML. This API is designed for developers integrating Azure TTS into applications requiring broad language and regional accent support. Key parameters include subscription key, region, voice name, language, and speech rate configuration.
- [Cartesia](https://pre.trtc.io/document/79685.md): Cartesia is a real-time voice AI text-to-speech (TTS) SDK designed for ultra-low latency streaming with natural-sounding output. It supports multilingual synthesis and voice mixing through its Sonic model, making it ideal for interactive voice agents requiring minimal end-to-end latency. This reference details the TTSConfig parameters used in the StartAIConversation API to configure Cartesia as the TTS engine, including model selection, API key, and voice ID.
- [ElevenLabs](https://pre.trtc.io/document/79686.md): ElevenLabs Text-to-Speech (TTS) configuration guide explains how to integrate ElevenLabs' advanced AI voice synthesis capabilities into Tencent RTC using the StartAIConversation API. It covers required parameters such as TTSType, Model, APIKey, and VoiceId, enabling developers to leverage professional voice cloning and multilingual voice options for realistic audio experiences.
- [Inworld](https://pre.trtc.io/document/79687.md): Inworld Text-to-Speech (TTS) is an AI-driven voice engine integrated with character personality and emotion systems, designed for gaming and virtual worlds. This API reference details how to configure Inworld TTS within the StartAIConversation API, including required parameters like APIKey, VoiceId, ModelId, and optional voice expressiveness control via Temperature.
- [Custom](https://pre.trtc.io/document/79688.md): This document explains how to integrate a custom Text-to-Speech (TTS) engine with the TRTC platform by implementing its custom streaming protocol. It is intended for developers who need to use proprietary or third-party TTS services not natively supported by TRTC. The guide covers configuration parameters for the StartAIConversation API, HTTP request headers, request and response formats, and error handling to enable full control over the TTS synthesis pipeline.
- [Intelligent Interruption Logic](https://pre.trtc.io/document/65319.md): Intelligent Interruption is a key feature of Tencent Real-Time Communication (Tencent RTC) AI conversation SDK that enhances real-time voice conversations by providing automatic and manual interruption modes. It helps users dynamically manage conversation flow by triggering interruptions based on audio duration or semantic detection, while also allowing manual control through SDK custom messages. This feature is designed for scenarios such as customer service, online education, and group meetings to improve interaction fluency and naturalness.
- [Context Management](https://pre.trtc.io/document/65318.md): Tencent Real-Time Communication (Tencent RTC) AI offers SDK features for managing context in large language models. This guide is designed for developers aiming to implement both basic and custom context management to improve response accuracy based on chat history. It details configuration using the History parameter for up to 50 conversation rounds and provides step-by-step instructions with sample code for integrating custom context management using OpenAI-compatible APIs and LangChain on backend services.
- [Customize TTS protocol](https://pre.trtc.io/document/65315.md): The Customize TTS protocol API provides a streaming text-to-speech service allowing clients to send text and receive synthesized audio in various formats such as WAV, MP3, or PCM. This API is designed for developers integrating Tencent RTC's TTS capabilities, supporting detailed request headers for task identification, authentication, and room context. It enables flexible audio output configuration including format, sample rate, and channel count. Responses include binary audio streams with corresponding Content-Type headers or JSON error messages for troubleshooting.
- [AI Conversation Status Callback](https://pre.trtc.io/document/68332.md): Tencent RTC Conversational AI status callback is a feature of the RTC Engine SDK that delivers real-time chatbot status updates such as Listening, Thinking, Speaking, and Interrupted. It is designed for developers to synchronize AI conversation states across all participants using JSON-formatted custom messages with a fixed cmdID. This capability enables UI feedback, interaction control, and conversation analysis to enhance user experience in real-time communication scenarios.
- [AI Conversation Subtitle Callback](https://pre.trtc.io/document/68333.md): Tencent RTC Conversational AI offers an SDK feature to display real-time subtitles synchronized at millisecond-level latency with conversation audio. This capability is designed for developers integrating AI conversation subtitles into their applications using custom JSON message formats and callback handling via the RTC Engine SDK. It enables flexible, accurate subtitle display and conversation segmentation for enhanced user experience.
- [AI Conversation Metric Callback](https://pre.trtc.io/document/68334.md): Tencent RTC Conversational AI metric callbacks SDK enables real-time monitoring of AI conversation performance by delivering detailed metrics such as STT latency, LLM network latency, TTS latency, and interruption counts via custom messages. This capability allows developers to track conversation duration and quality on client platforms using the RTC Engine SDK's Receive Custom Messages feature.
- [AI Conversation Error Callback](https://pre.trtc.io/document/68335.md): The AI Conversation Error Callback feature of Tencent RTC enables clients to monitor conversational AI performance by receiving real-time error notifications via custom messages. This API reference details the structure of error callback messages, including metric types such as ASR, LLM, and TTS errors, and provides the associated error codes and descriptions to facilitate troubleshooting and quality monitoring in AI conversations.
- [AI Conversation Server Webhooks](https://pre.trtc.io/document/68331.md): The AI Conversation Server Callbacks API documentation details the event notifications sent by Tencent Cloud's Conversational AI service. It explains how to configure callback URLs, the message format, event types, and parameters for receiving AI task and session status updates via HTTP/HTTPS POST requests. This guide is intended for developers integrating AI conversation event callbacks with Tencent RTC to implement custom logic and monitor AI service events reliably.
- [StartAIConversation](https://pre.trtc.io/document/64963.md): The StartAIConversation API is part of the Tencent Real-Time Communication (TRTC) SDK, enabling developers to initiate AI-driven conversations within TRTC rooms. It allows an AI Channel Robot to join rooms and interact with specified members using integrated speech-to-text, third-party large language models (LLM), and text-to-speech (TTS) services. This API supports up to 50 requests per second and is designed for scenarios such as smart customer service and AI language teaching. Supported regions include Guangzhou, Singapore, Tokyo, Ashburn, and Silicon Valley. Key input parameters include SdkAppId, RoomId, AgentConfig, LLMConfig, and TTSConfig, facilitating flexible AI dialogue configurations.
- [UpdateAIConversation](https://pre.trtc.io/document/64962.md): The UpdateAIConversation API is part of the Tencent RTC TRTC API suite, designed to update AI conversation task parameters. It supports specific regions including ap-guangzhou, ap-singapore, ap-tokyo, and na-siliconvalley, and allows up to 20 requests per second. This API accepts parameters such as TaskId, WelcomeMessage, InterruptMode, LLMConfig, and TTSConfig, and returns a unique RequestId for each call. Developer resources include Tencent Cloud SDK 3.0 for multiple languages and Tencent Cloud CLI 3.0. The documentation also details error codes related to the API's business logic for troubleshooting.
- [ControlAIConversation](https://pre.trtc.io/document/64965.md): ControlAIConversation API is a Tencent RTC service API designed for server-side robot control. It enables developers to send commands such as ServerPushText to broadcast text through AI robots or InvokeLLM to trigger large model dialogues. The API supports multiple regions including Guangzhou, Singapore, Tokyo, Ashburn, and Silicon Valley, and allows up to 50 requests per second. Comprehensive SDKs and CLI tools are available to facilitate integration across various programming languages.
- [DescribeAIConversation](https://pre.trtc.io/document/64964.md): DescribeAIConversation API is part of Tencent RTC SDK that allows querying the status of AI conversation tasks. It supports specific regions including Guangzhou, Singapore, Tokyo, and Silicon Valley, and handles up to 20 requests per second. The API returns detailed task information such as start time, status, task ID, and session ID. Developers can utilize Tencent Cloud SDK 3.0 in various programming languages and the Command Line Interface to integrate and manage AI conversation tasks efficiently.
- [StopAIConversation](https://pre.trtc.io/document/65296.md): The StopAIConversation API from Tencent RTC allows developers to terminate AI conversation tasks by specifying a unique TaskId. This API supports up to 20 requests per second and operates in regions including Guangzhou, Singapore, Tokyo, and Silicon Valley. It returns a unique RequestId for each request to assist in troubleshooting. The API is accessible via Tencent Cloud SDK 3.0 in multiple programming languages and includes documented error codes for task validation.
- [Emotional Companionship](https://pre.trtc.io/document/64660.md): Tencent Real-Time Communication (Tencent RTC) offers an SDK platform for AI emotional companionship, enabling developers to create immersive real-time voice and video interactions. It supports diverse scenarios such as virtual companions, role-playing, and AI avatars, providing personalized customization, emotional understanding, and seamless content creation. Tencent RTC ensures low-latency, high-quality communication and high concurrency to deliver smooth, realistic, and emotionally adaptive companionship experiences for users.
- [Intelligent Voice Customer Service](https://pre.trtc.io/document/75308.md): AI Customer Service is an intelligent voice customer service solution that leverages Tencent RTC's real-time communication technology combined with AI, ASR, LLM, STT, and TTS modules. It enables businesses to implement automated, coherent, and natural voice interactions with customers, reducing latency and improving user satisfaction. This guide covers the solution architecture, prerequisites, integration steps, and advanced features such as custom messaging, digital human video service, and context management to help developers deploy and optimize intelligent customer service systems.
- [AI Interview](https://pre.trtc.io/document/75309.md): AI Interview SDK is a comprehensive solution that integrates artificial intelligence with Tencent RTC's real-time audio/video communication capabilities to automate online interviews. Designed for developers and enterprises, it enables rapid deployment on iOS, Android, Windows, Mac, Web, and Mini Programs. This quickstart guide covers essential steps including importing the RTC Engine SDK, entering rooms, publishing audio streams, and configuring conversational AI with large language models and text-to-speech services to deliver natural and efficient interview experiences.

## Speech To Text

- [Server-Side Integration of AI Transcription and Translation](https://pre.trtc.io/document/79873.md): Tencent RTC AI Transcription and Translation SDK enables real-time speech-to-text conversion and multi-language translation in TRTC rooms. This guide explains how to integrate these features on the server side by importing the TRTC SDK, joining rooms, and initiating transcription tasks via RESTful API calls. It covers prerequisites, task creation, callback handling, and best practices to enable seamless real-time transcription and translation for users speaking the same or different languages.
- [SDK Integration for AI Transcription/Translation](https://pre.trtc.io/document/79355.md): The TRTC SDK provides an AI real-time transcription and translation SDK for iOS, Android, Windows, macOS, and Web clients. It enables developers to quickly integrate speech-to-text and multi-language translation features on the client side using the AITranscriberManager interface. This guide covers the integration prerequisites, obtaining the AITranscriberManager instance, setting up event listeners, and managing transcription tasks to streamline real-time audio conversion and translation without backend dependencies.
- [Integrating AI Transcription and Translation into Video Calls](https://pre.trtc.io/document/80096.md): TUICallKit, part of the TRTC Audio/Video Call SDK, enables developers to quickly build video call applications with AI-powered real-time transcription and translation. Supporting multiple platforms such as iOS, Android, Web, Flutter, and more, TUICallKit offers a seamless integration process in just a few steps. This solution guides users through activating the service, integrating the SDK, and enabling AI transcription and translation features, helping reduce development time and enhance call accessibility with live captions and multilingual support.
- [Multi-party Meeting Integration with AI Transcription and Translation](https://pre.trtc.io/document/80346.md): TUIRoomKit is a multi-party conference SDK designed for scenarios such as enterprise meetings, webinars, and online education, supporting platforms including Web and Android. This solution guide explains how to quickly integrate TUIRoomKit and enable AI real-time subtitles and meeting notes to enhance conference interactions. It covers activation steps, billing details for both the SDK and AI transcription services, and platform-specific integration instructions to help developers build applications with live transcription capabilities efficiently.
- [AI Transcription and Translation Callback](https://pre.trtc.io/document/80094.md): This document provides the API reference for AI Transcription and Translation Callback in TRTC. It explains how to configure Event Webhook notifications, the message format, event types, and parameters for backend integration. The API supports secure callback delivery with signature verification and retry mechanisms to ensure reliable notification of transcription and translation events.
- [Real-time Speech Recognition](https://pre.trtc.io/document/80841.md): The Real-time Speech Recognition API is a WebSocket-based service that enables synchronous transcription of audio streams in multiple languages. Designed for general-purpose use, it supports PCM audio input at 8kHz or 16kHz sample rates and returns JSON-formatted recognition results. This API is suitable for developers integrating real-time speech recognition into applications, offering detailed call flow stages, concurrency management, and authentication methods using SDKAppId and UserSig.
- [One-Sentence Recognition](https://pre.trtc.io/document/80843.md): The One-Sentence Recognition API is a speech-to-text service designed to transcribe short audio clips up to 60 seconds in duration. It supports multiple audio formats such as wav, pcm, ogg-opus, mp3, and m4a, and offers recognition in over 30 languages. This API is intended for developers integrating speech recognition capabilities into applications, providing key features like HTTP POST request handling, JSON response formatting, and concurrency management. The API accepts audio data either via direct upload or accessible URLs, ensuring flexible integration options.
- [Audio File Recognition](https://pre.trtc.io/document/80844.md): The Audio File Recognition API is a Tencent RTC service that provides speech recognition capabilities for long audio and video files across multiple languages. It supports various audio and video formats, allowing users to submit files via URLs or local uploads. The API returns recognition results in JSON format and supports retrieval through webhooks or polling. It enforces concurrency limits and requires authentication with TRTC credentials for secure access.

## Text To Speech

- [Other Scenarios](https://pre.trtc.io/document/81360.md): Conversational Text To Speech is a speech synthesis SDK built on Tencent Real-Time Communication (TRTC) designed for real-time conversational applications. It provides ultra-low latency streaming and non-streaming APIs, high-quality natural voices in multiple languages including Chinese, English, Japanese, and Cantonese, and supports rapid voice cloning for custom voice creation. This service enables developers to integrate realistic AI speech for scenarios such as live dubbing, audiobooks, and customer service.
- [Voice List](https://pre.trtc.io/document/81361.md): The Premium Voice Library is a comprehensive voice SDK offering multilingual voices in Chinese, English, Japanese, and Cantonese. It is designed for developers and content creators to enhance character dubbing, audiobooks, customer service, and announcement scenarios by providing diverse voice tones and styles tailored to specific use cases.