Gemini-SQL2: text-to-SQL with Gemini 3.1 Pro, SOTA on BIRD

Melvin Vivas · X post · 2026-06-13 · Open on X

Topics: AI Dev Tools & Productivity, Evaluation (Evals) & Testing, Industry Trends & Job Market · Level: intermediate

Summary

The creator quotes Google's announcement of Gemini-SQL2, a text-to-SQL feature powered by Gemini 3.1 Pro. The announcement says it reached state-of-the-art results on the BIRD benchmark by turning natural language into ready-to-run SQL. It shows that natural-language-to-SQL is becoming practical.

Key points

• Gemini-SQL2 turns natural-language questions into SQL queries that can be run directly.
• It is powered by Gemini 3.1 Pro.
• It is reported to reach state-of-the-art results on the BIRD text-to-SQL benchmark.
• BIRD is a competitive benchmark for checking how well models turn questions into SQL.
• The creator's take: you may no longer need to write SQL by hand.

Resources mentioned

• Gemini-SQL2 · tool · check price
  Google's text-to-SQL feature that turns natural language into SQL queries ready to run.
• Gemini 3.1 Pro · tool · deepmind.google · check price
  Google's Gemini model that powers Gemini-SQL2.
• BIRD benchmark · dataset · bird-bench.github.io · free
  A large text-to-SQL benchmark that tests how well models turn questions into SQL on real databases.

Try this

• Build a natural-language-to-SQL assistant and test it on the BIRD benchmark.