BigQuery provides extensive AI/ML capabilities with core Gemini features available free and advanced features in Enterprise tier. Free features include SQL code assist (Gemini suggestions in console), Data Canvas (visual data exploration), and basic data preparation assistance. BigQuery ML enables model training and inference directly in SQL: built-in models (linear/logistic regression, k-means clustering, ARIMA forecasting, TimesFM for time series) run at standard query pricing. ML.GENERATE_TEXT enables direct Gemini calls within SQL queries for text analysis, summarization, and classification. External model support connects to Vertex AI endpoints for custom models. Duet AI features (renamed to Gemini) assist with query writing, optimization suggestions, and error explanation. BigQuery Studio provides notebook-style environment with Python, SQL, and Spark in unified interface. No additional licensing for core AI features—usage billed at standard BigQuery rates plus model-specific costs for ML.GENERATE_TEXT calls. Page should cover: free vs paid feature matrix, BigQuery ML model types and use cases, ML.GENERATE_TEXT examples and pricing, BigQuery Studio setup, comparison with Snowflake Cortex, and cost optimization strategies.