Agora enables more natural voice conversations with AI, thanks to low-latency responses and real-time interruption handling. Agora’s built-in background noise suppression, echo cancelation, and selective attention locking allow AI to hear the user clearly in any environment. Agora’s global real-time network ensures connectivity and performance in any location.

Agora's Conversational AI Engine offers support for a wide range of large language models (LLMs), including:

• OpenAI
• OpenAI Realtime API
• Azure OpenAI
• Google Gemini
• Google Vertex AI
• Anthropic Claude
• Dify
• Custom LLM

Review our documentation on connecting LLMs here: https://docs.agora.io/en/conversational-ai/models/llm/overview

Agora’s Conversational AI Engine currently supports the following ASR providers:

• ARES (default)  
• Microsoft Azure
• Deepgram

Review our documentation on connecting ASR models here: https://docs.agora.io/en/conversational-ai/models/asr/overview

Agora’s Conversational AI Engine currently supports the following TTS providers:

• Microsoft Azure
• ElevenLabs
• Cartesia (Beta)
• OpenAI (Beta)
• Hume AI (Beta)

Review our documentation on connecting TTS models here: https://docs.agora.io/en/conversational-ai/models/tts/overview

Agora’s Conversational AI Engine currently supports the following AI avatar providers:

• Akool (Beta)
• HeyGen (Alpha)

Review our documentation on connecting avatar providers here: https://docs.agora.io/en/conversational-ai/models/avatar/overview

To implement a voice AI agent, you need to connect an LLM and a text-to-speech service to Agora’s Conversational AI Engine. This enables full customization of the experience, with the LLM and voice of your choice.

The chained or cascade model refers to the processing flow of the user’s voice being processed by automatic speech recognition (ASR) technology that converts speech to text, then that text being processed by the LLM, then the LLM’s response being processed by text-to-speech technology and ultimately outputting the AI agent’s voice response.

No, Agora’s Conversational AI Engine requires an existing AI model or LLM. The Engine enables customized voice interaction with the LLM but is not capable of creating or training an LLM.

Agora Voice Calling is a real-time voice API that lets developers embed high-quality, ultra-low latency voice chat into any application. It supports one-to-one calls, group voice chat, and large-scale audio rooms across devices and platforms.

Agora Voice Calling supports Android, iOS, Web, Windows, Electron, Flutter, React Native, Unity, and Unreal Engine. This allows teams to build consistent voice experiences across mobile, web, desktop, and immersive environments.

Agora uses a 48 kHz sampling rate with full-bandwidth audio capture and intelligent routing over its global real-time network. This minimizes latency, jitter, and packet loss to deliver clear, stable voice calls—even on unstable networks.

Yes. Agora Voice Calling includes AI-powered features such as Noise Suppression, Real-Time Speech to Text, and seamless integration with large language models and text-to-speech engines to enable intelligent, voice-driven experiences.

Yes. Agora supports flexible voice recording in the cloud or on premises. Developers control audio formats, storage locations, and recording quality to support playback, analytics, moderation, or compliance needs.

3D Spatial Audio simulates real-world sound positioning, making conversations feel more immersive and natural. It’s commonly used in gaming, social audio rooms, virtual workspaces, and metaverse-style experiences.

You can integrate Agora Voice Calling within hours using SDKs, documentation, and sample apps. For teams that want to move faster, Agora App Builder offers a no-code option to deploy voice chat without custom development.

Agora Voice Calling is ideal for education platforms, multiplayer games, social apps, collaboration tools, live shopping, customer engagement, and IoT devices—any use case that requires reliable, real-time voice communication at global scale.

Agora Video Calling is a real-time video API that lets developers embed high-quality, low-latency video calls into web, mobile, and native applications. It supports everything from 1:1 calls to large-scale video experiences with full customization.

Agora Video Calling supports Android, iOS, Web, Windows, Electron, Flutter, React Native, Unity, and Unreal Engine—making it easy to deliver consistent video experiences across devices and operating systems.

Agora uses intelligent routing and adaptive video optimization to reduce jitter, lag, and packet loss. The platform dynamically adjusts video quality in real time to maintain smooth, uninterrupted calls—even on slow or unstable networks.

Agora supports advanced collaboration features such as screen sharing, interactive whiteboards, multi-user video layouts, and real-time messaging. These features make it well suited for meetings, education, telehealth, and collaborative work apps.

Yes. Agora provides flexible video call recording options, allowing you to record securely to the cloud or on local servers. Developers control video format, resolution, storage location, and access permissions to meet compliance and operational needs.

Yes. Agora supports multi-track audio and video, making it possible to publish multiple camera feeds or microphone streams within a single session. This is ideal for live production workflows, virtual events, and advanced conferencing scenarios.

You can ship a video calling app within hours using Agora SDKs, documentation, and sample apps. For even faster deployment, Agora App Builder provides a no-code option to launch video, voice, and live streaming features without custom development.

Agora Video Calling is ideal for education, remote work, gaming, social apps, live shopping, and telehealth. Any application that requires scalable, real-time video communication with global reach and low latency can benefit from Agora’s platform.

Agora Real-Time Chat is a customizable chat SDK that lets developers add secure, scalable messaging to real-time video, voice, and live streaming applications. It supports one-to-one messaging, group chat, and large community channels.

Agora’s Chat SDK supports Android, iOS, Web, Windows, Flutter, React Native, and Unity, making it easy to deliver consistent messaging experiences across mobile, desktop, and cross-platform apps.

Agora Chat supports rich media messaging including emojis, images, files, GPS locations, structured messages, and voice notes. Core messaging features also include offline messaging, message recall and deletion, read receipts, typing indicators, presence, and push notifications.

Agora Chat uses TLS/SSL encryption for data in transit and encrypted file storage to protect user data. The platform also supports privacy compliance features such as user data deletion and secure message handling.

Yes. Agora Chat includes built-in content moderation to help filter profanity, offensive language, and inappropriate images or text. Developers can also integrate third-party moderation tools for additional control.

Yes. Agora Chat supports multilingual message translation with automatic, on-demand, or push-based translation options, enabling users to communicate in their preferred language.

Developers can launch a chat experience within hours using Agora SDKs, documentation, and sample apps. For faster implementation, Agora UI Kit provides a low-code option to add messaging with minimal development effort.

Agora Real-Time Chat is ideal for education platforms, gaming communities, social apps, collaboration tools, live commerce, and telehealth—any application that requires reliable, secure, and engaging real-time messaging.

Agora Real-Time Speech to Text is a cloud-based live transcription and subtitling service that converts real-time audio into accurate text for live audio and video applications. It enables captions, transcripts, and AI-powered workflows without impacting real-time performance.

Agora’s cloud-based transcription processes audio streams in real time and converts speech into text with low latency. Transcripts can be delivered as live captions to participants, stored for later review, or exported for downstream processing.

Yes. Real-time transcripts can be integrated with large language models to generate summaries, meeting notes, action items, feedback, or translations. Transcripts can also be exported as .vtt files for seamless LLM processing without affecting RTC performance.

Yes. Agora supports real-time speaker recognition and labeling for up to three simultaneous speakers. Each speaker can be transcribed separately, improving accuracy in conversations with interruptions or overlapping dialogue.

Agora supports all major languages and regional dialects. Each channel can transcribe up to two languages simultaneously, making it ideal for multilingual meetings, events, and global applications.

Yes. Agora supports transcription for cloud-recorded audio and video, enabling closed captions (CC) during playback and searchable transcripts for reviewing important discussion points.

Agora uses advanced AI techniques to reduce silence, lower Word Error Rate (WER), and maintain accuracy even with accents, overlapping speech, poor audio quality, or unstable networks. The solution scales from one-to-one sessions to millions of participants with consistent accuracy.

Yes. Agora is ISO and SOC 2 certified and supports compliance with GDPR, CCPA, and HIPAA. Live captions and transcripts can be encrypted using the same security mechanisms as Agora’s real-time audio and video streams.

Agora Recording is an extension that allows developers to record audio streams, video streams, interactive content, and web pages for archive, review, compliance, or redistribution. It supports cloud, on-premises, and webpage recording options.

Agora Recording can capture audio, video, screen content, whiteboards, chat messages, and live streaming elements. You can record single streams or multiple streams separately, making it easy to edit, combine, or repurpose content later.

Single-stream recording combines audio, video, and content into one synchronized file. Multi-stream recording captures each audio, video, or content stream separately, giving you greater flexibility for post-production, analysis, or moderation workflows.

Recordings can be stored in the cloud or on-premises, depending on your deployment needs. Agora supports third-party cloud storage providers such as Amazon S3, Microsoft Azure, Google Cloud, Alibaba Cloud, Tencent Cloud, and others.

Yes. Agora Recording supports screenshots for moderation, customizable capture intervals, digital watermarks, and content moderation tools. These features help enforce community guidelines, protect intellectual property, and meet regulatory or organizational requirements.

Agora Recording is built with enterprise-grade security, including end-to-end encryption for calls, transmission, and storage. It supports globally distributed clusters, automatic backups, proxy services, and LAN deployment to meet strict data security and privacy needs.

Developers can integrate Agora Recording in as little as 30 minutes using RESTful APIs. The service is designed to be easy to embed, test, and deploy, with automatic uploading and backup to ensure recordings are not lost.

Agora Recording is ideal for virtual events and webinars, large-scale live streaming, customer service quality assurance, education and online classes, and telehealth consultations—any scenario where capturing, reviewing, or distributing real-time interactions is essential.