LiveKit Full Beginner Tutorial (2025)

𝘓𝘪𝘷𝘦𝘒𝘪𝘵 is a robust open-source platform for creating voice AI and multimodal AI agents, trusted by top industry players like eBay and OpenAI.⁣ ⁣ In this session, you’ll learn:⁣ ⁣ 𝐇𝐨𝐰 𝐭𝐨 𝐒𝐞𝐭 𝐔𝐩 𝐘𝐨𝐮𝐫 𝐀𝐈 𝐕𝐨𝐢𝐜𝐞 𝐀𝐠𝐞𝐧𝐭:⁣ - Leverage LiveKit Cloud, which provides 5,000 free minutes per month.⁣ - Utilize Gemini 2.0 Flash Experimental as the LLM (free to use).⁣ - Integrate Deepgram for TTS (Text-to-Speech) and STT (Speech-to-Text) with a $200 free credit, making your voice agent completely free to use.⁣ ⁣ 𝐔𝐩𝐠𝐫𝐚𝐝𝐞 𝐕𝐨𝐢𝐜𝐞 𝐐𝐮𝐚𝐥𝐢𝐭𝐲 𝐚𝐧𝐝 𝐋𝐚𝐭𝐞𝐧𝐜𝐲:⁣ Replace Deepgram TTS with ElevenLabs for better voice quality and lower latency.⁣ ⁣ 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐡𝐞 𝐓𝐮𝐫𝐧 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐌𝐨𝐝𝐞𝐥:⁣ Add LiveKit’s turn-detection model to ensure the AI responds naturally after the user finishes speaking, enabling seamless interactions without interruptions.⁣ ⁣ ✨ 𝐍𝐨 𝐜𝐨𝐝𝐢𝐧𝐠 𝐞𝐱𝐩𝐞𝐫𝐢𝐞𝐧𝐜𝐞? 𝐍𝐨 𝐰𝐨𝐫𝐫𝐢𝐞𝐬!⁣ I’ll guide you through every step, making it easy to follow and implement on your own.⁣ VAPI ($10 free credits): https://vapi.ai/?via=sanava Retell AI (60 free call mins): https://dashboard.retellai.com//?ref=sanava LiveKit: https://livekit.io/ Get 20 Euros of free Hetzner credits: https://hetzner.cloud/?ref=bsofzNyuiHSb FREE RESOURCES : https://sanava-ai.com/resource-hub/ 𝐉𝐨𝐢𝐧 𝐨𝐮𝐫 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 𝐜𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐢𝐞𝐬: Voice AI Discord (over 900 members): https://discord.gg/6yck2C6RaD LinkedIn Voice AI Group: https://www.linkedin.com/groups/10008267/ Facebook Voice AI Group: https://www.facebook.com/groups/908425741232407 If you need a Voice AI solutions for your business, reach out to me at [email protected] For any questions: [email protected] ☕ 𝐁𝐮𝐲 𝐌𝐞 𝐀 𝐂𝐨𝐟𝐟𝐞𝐞: https://www.paypal.com/paypalme/noursiakiroglou Timestamps: 00:00 Who is this session for? 01:09 Retell AI, Vapi, LiveKit Comparison 11:48 Today's Session Outline 15:28 Step 1: Download and Set up a LiveKit Voice AI Agent Template 29:15 Step 2: Change LLM to Gemini and TTS to Deepgram 33:00 Generate Google API Keys 43:20 Troubleshooting 50:53 Free Voice AI Demo 51:47 Step 3: Add Turn Detector 58:00 Turn Detector Demo 1:00:00 Change TTS to Elevenlabs 1:11:17 Gemini Voice Agent With Elevenlabs TTS Demo