Google Launches Gemini Live Agent Challenge

Got story updates? Submit your updates here. ›

Google is challenging developers to build the next generation of multimodal AI agents that can see, hear, speak, and create as part of the Gemini Live Agent Challenge. Participants can choose from three categories - The Live Agent, The Creative Storyteller, or The UI Navigator - and use Gemini models and the Agent Development Kit to create innovative solutions. The grand prize winner will receive $25,000, a trip to Google Cloud Next '26 in Las Vegas, and the opportunity to present on stage.

Why it matters

This challenge represents Google's push towards more immersive, real-time AI experiences that go beyond the traditional text-based interactions. By encouraging the development of multimodal AI agents, Google is aiming to advance the state of the art in conversational AI, creative tools, and visual navigation.

The details

Developers must use a Gemini model, such as Gemini 3 or Nano Banana, and the Gen AI SDK or Agent Development Kit (ADK) to build their solutions. They must also leverage at least one Google Cloud service, such as Firestore, CloudSQL, Cloud Run, or Vertex AI. Submissions are open until March 16, 2026.

• The Gemini Live Agent Challenge was announced on March 7, 2026.
• Submissions for the challenge are open until March 16, 2026.

What’s next

Developers interested in participating in the Gemini Live Agent Challenge can register on the hackathon website, watch the kickoff video, and review the official rules.