Google AI for game developers- Google Developers Blog
May 9, 2025Jetha Chan, Ju-yeong Ji, Ishan Sharma, Avneet Singh

Summary
This post introduces Gemma 3 and Gemini 2.5 models, detailing their capabilities for game development, including multimodal input, extended context, function calling, and multi-language support. It announces the Gemma Unity plugin, built on Gemma.cpp for on-device inference, and the 'Gemma Journey' sample game demonstrating prompt-driven NPC dialogue. The 'Home Run: Gemini Coach Edition' demo is presented as an example of Gemini 2.0 Flash for real-time coaching in a mobile game. Google Cloud services like Vertex AI and Agones are recommended for scaling game development with generative AI.
More from the Google Developer Platform trail
- Autonomous LLM post-training with Tunix on TPUs- Google Developers Blog
- HeyGen x Google Cloud: Bringing Avatar IV to TPUs- Google Developers Blog
- Run Ray on TPU, Part 2: Ray AI libraries- Google Developers Blog
- Scaling Agentic RL: High-Throughput Agentic Training with Tunix- Google Developers Blog
- Run Ray on TPU, Part 1: The foundations- Google Developers Blog