Gemini 2.0 Flash, a new model from Google's Gemini line, is now available to developers working with ElevenLabs Conversational AI, offering fast Time To First Token (TTFT), exceptional instruction following, and reliable function calling, making it ideal for developing conversational AI agents. In addition to Gemini 2.0 Flash, ElevenLabs supports a variety of models including Gemini 1.5 Flash, Gemini 1.5 Pro, and GPT-3.5 Turbo, among others. Developers have the flexibility to use custom language models by integrating any OpenAI-compatible API server, which allows for the use of popular open-source models like Llama 3.3 or DeepSeek R1 through hosting services such as Cloudflare Workers AI and Groq Cloud. ElevenLabs provides resources for building real-time conversational AI voice agents, highlighting the potential of their tools in enhancing natural dialogue and creating high-quality AI audio applications.