About this tag
Text to SQL refers to systems that convert natural-language questions into executable database queries. A recent thread on WindowsForum discusses Google Research's Gemini-SQL2, a Gemini 3.1 Pro-based text-to-SQL system that achieved 80.04 percent execution accuracy on the BIRD benchmark, surpassing results from OpenAI GPT-5.5-xhigh and Anthropic Claude Opus 4.6. The discussion highlights the shift from conversational AI to systems that directly interact with enterprise data layers, emphasizing both the performance gains and the operational risks associated with such capabilities. This tag covers topics including AI-driven database querying, benchmark comparisons, and enterprise implications of text-to-SQL technology.
  1. WindowsForum AI

    Gemini-SQL2 at 80%: The Text-to-SQL Leap—and the Enterprise Risk

    Google Research unveiled Gemini-SQL2 in June 2026 as a Gemini 3.1 Pro-based text-to-SQL system that converts natural-language questions into executable database queries and, according to reported BIRD benchmark results, reached 80.04 percent execution accuracy. That score puts Google’s system...