This workshop teaches how to build advanced AI applications with the Google Gemini 2.5 model family, focusing on practical engineering skills for building agentic AI systems.
Note
The notebooks include the workshops and learning exercises. You'll find solutions in the solutions folder.
Prerequisites: You need an API key from Google AI Studio. Everything can be done on the free tier. Install the google-genai package with pip install -U -q "google-genai".
Acknowledgment: This workshop and structure is inspired by Patrick Loeber's excellent Gemini workshop. Check out his work for additional learning resources.
Start in colab: 
- Google AI Studio setup and API key configuration
- Installing the Python SDK (
google-genai)
- Basic text generation and streaming responses
- Token counting and cost management
- Multi-turn chat conversations
- System instructions and model configuration
- Long context handling and file uploads
- Image understanding and analysis (single and multiple images)
- Audio processing (transcription, analysis, summarization)
- Video understanding (summarization, transcription)
- Document processing (PDFs, structured data extraction)
- Text-to-speech generation & Image generation
- Structured outputs with Pydantic schemas
- Function calling and external API integration
- Native tools (code execution, Google Search, grounding)
- Automatic function calling capabilities
- Introduction to Model Context Protocol
- Working with stdio and HTTP MCP servers
- Building interactive chat agents with MCP