Your
agent can listen.Now give it a face.
Start building Deepgram listens. Your chosen voice responds. Cara-4.1 Express brings the conversation to life with natural facial expression and lip sync, streamed through the Anam JavaScript SDK.
See it for yourself
KEEP EXPLORING
Build more with Anam + Deepgram
DEEPGRAM SPEECH-TO-TEXT · SERVER-SIDE
Build with Deepgram and Anam.
Choose Deepgram as the speech-to-text provider in your Anam persona. Your server creates a session token and the Anam JavaScript SDK streams the conversation to your app. Your selected LLM and text-to-speech voice provide the replies.
// Inside POST /api/anam-session. Keep ANAM_API_KEY on the server.
// avatarId, voiceId, llmId, and systemPrompt come from server config.
// Use Anam catalog IDs for your chosen avatar, TTS voice, and LLM.
const response = await fetch("https://api.anam.ai/v1/auth/session-token", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.ANAM_API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
personaConfig: {
name: "Deepgram STT avatar",
avatarId,
avatarModel: "cara-4.1-express",
voiceId,
llmId,
systemPrompt,
languageCode: "en",
voiceDetectionOptions: {
sttProvider: "deepgram",
},
},
}),
});
if (!response.ok) throw new Error("Could not create Anam session");
const { sessionToken } = await response.json();
// Return { sessionToken } as JSON from your /api/anam-session route.
Cara-4.1 Express
Highest-rated in a blind benchmark against Tavus, HeyGen and LemonSlice.
Plus, fastest connection time and fastest reply latency.
See Avatar Benchmark (October 2026)Add a face to your Deepgram STT agent.
Bring Cara-4.1 Express to your app with natural facial expression and lip sync that follows your agent's voice.