ElevenLabs releases Unity SDK for voice agents
ElevenLabs has shipped a v0.1.0 Unity SDK for its ElevenAgents platform, letting game developers add spoken conversations with in-game characters through the company's Conversation API.
ElevenLabs announced the ElevenAgents Unity SDK, which connects a Unity scene to an ElevenLabs voice agent. The SDK takes input from the player's microphone, plays the agent's voice through a Unity AudioSource and runs the session through the Conversation API.
The same code runs in the Editor, desktop builds and WebGL. Support for iOS, Android and XR is in beta. The SDK chooses the transport for each build target. Each agent, including its prompt and voice, is configured in the ElevenLabs dashboard and referenced by ID.
Developers can attach the voice to a character's own AudioSource and use GetOutputVolume() to drive mouth animation. Dynamic variables let several characters share one agent while passing their own values, such as mood. Per-session Overrides can replace settings like the first message or the voice, but each override must first be enabled in the agent's Security settings. SendContextualUpdate tells the agent about game events mid-conversation without adding a player turn. Client tools let the agent trigger game code through registered handlers. Events such as transcripts and interruptions fire on Unity's main thread.
The release has platform caveats. WebGL requires the WebAssembly.Table setting, and some AudioSource features, including mixer routing and doppler, don't work there. WebRTC sessions ignore custom output sources.
Developers install the package from a git URL pinned to v0.1.0. ElevenLabs warns that agent authentication should be turned back on before shipping. Builds should then connect with short-lived signed URLs from the developer's server, and API keys should never be embedded in a game build. The company did not mention pricing or any production titles using the SDK.