Skip to main content

About VideoSDK

What is VideoSDK?​

VideoSDK is a real-time communication platform for developers. It gives you the media infrastructure, client SDKs, and APIs to add live audio, video, live streaming, and AI voice agents to your own product, without operating your own media servers.

You bring the product and the interface. VideoSDK handles media routing, scaling, recording, and the platform work underneath: signaling, network adaptation, geo-distributed media servers, and access control.

About WebRTC​

WebRTC is the browser and mobile standard for sending audio and video between peers in real time. It ships with codecs, encryption, and network adaptation tuned for latency instead of throughput, which is what makes sub-second conversation possible.

WebRTC alone is only a transport. A production call also needs a media server to route streams between many participants, token-based authentication, session state, recording, and observability. That is the layer VideoSDK provides.

Why use VideoSDK?​

  • Ship faster. A working audio and video call is a quickstart away in your framework of choice, instead of weeks of media plumbing.
  • One platform, every client. Native SDKs for web, mobile, desktop, game engines, and embedded devices share the same room model and the same tokens.
  • Managed infrastructure. Global media servers, scaling, and reliability are operated for you.
  • Composable. Rooms, recordings, transcription, live streams, AI agents, and phone calls are pieces you can combine in one session.
  • Server-authoritative security. Access tokens are minted on your backend with the permissions you decide, so clients never hold your API secret.

How does VideoSDK work?​

VideoSDK Cloud​

The media backbone. It runs the selective forwarding units that receive each publisher's tracks and forward them to subscribers, manages room and participant state, and handles recording, live streaming, and transcription jobs.

Client SDKs​

Native libraries that join a room, publish and subscribe to tracks, and expose participant events. Available for JavaScript, React, Angular, React Native, Android, iOS, Flutter, Unity, Python, C++, and IoT devices. Prebuilt SDK gives you a complete, themeable meeting UI when you do not want to build one.

AI Agent SDK​

A Python framework for real-time voice agents. An agent joins a room as a participant, streams audio in and out with your chosen speech and language models, and can be pointed at a phone number for telephony use cases.

Server SDK and REST APIs​

Your backend mints access tokens, creates and validates rooms, and controls recording, HLS, and RTMP. Use it to enforce who can join, who can publish, and what each participant is allowed to do.

Telephony​

SIP gateways connect external providers such as Twilio, Vonage, Plivo, or Telnyx to VideoSDK rooms. Routing rules decide how inbound numbers map to rooms and how outbound calls are placed from a session.

Where to go next​

  • Video and audio calling quickstart to get a call running.
  • Concept and architecture for the room, participant, and token model.
  • Interactive live streaming for host, audience, and HLS.
  • AI Voice Agents to build a real-time voice agent.
  • Telephony to connect phone numbers and SIP.
  • Server SDK for tokens, rooms, and recordings.
  • Code samples for full open source apps you can run today.

Pick your SDK in the sidebar to start, or sign up for free credits and an API key.