# SQL-Zero: Self-Evolving Text-to-SQL

arXiv: https://arxiv.org/abs/2609.04697 · HTML: https://mlsanity.com/paper/2609.04697/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Synthetic reasoning data and bootstrapping, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 18 of its 25 references that are in the map.
- First posted on arXiv 4 Sep 2026.

## Placed in

- Synthetic data and self-improvement for LLMs / [Synthetic reasoning data and bootstrapping](https://mlsanity.com/synth/synthetic-reasoning-data-and-bootstrapping/)

## Builds on (most cited first, 18 of 18)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task](https://mlsanity.com/paper/1809.08887/) (2018)
- [Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs](https://mlsanity.com/paper/2305.03111/) (2023)
- [Absolute Zero: Reinforced Self-play Reasoning with Zero Data](https://mlsanity.com/paper/2505.03335/) (2025)
- [R-Zero: Self-Evolving Reasoning LLM from Zero Data](https://mlsanity.com/paper/2508.05004/) (2025)
- [Towards Robustness of Text-to-SQL Models against Synonym Substitution](https://mlsanity.com/paper/2106.01065/) (2021)
- [OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale](https://mlsanity.com/paper/2503.02240/) (2025)
- [SQL-R1: Training Natural Language to SQL Reasoning Model By Reinforcement Learning](https://mlsanity.com/paper/2504.08600/) (2025)
- [Synthesizing Text-to-SQL Data from Weak and Strong LLMs](https://mlsanity.com/paper/2408.03256/) (2024)
- [Dr. Zero: Self-Evolving Search Agents without Training Data](https://mlsanity.com/paper/2601.07055/) (2026)
- [Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Data](https://mlsanity.com/paper/2602.21320/) (2026)
- [CSC-SQL: Corrective Self-Consistency in Text-to-SQL via Reinforcement Learning](https://mlsanity.com/paper/2505.13271/) (2025)
- [Think2SQL: Reinforce LLM Reasoning Capabilities for Text2SQL](https://mlsanity.com/paper/2504.15077/) (2025)
- [ExeSQL: Self-Taught Text-to-SQL Models with Execution-Driven Bootstrapping for SQL Dialects](https://mlsanity.com/paper/2505.17231/) (2025)
- [SPFT-SQL: Enhancing Large Language Model for Text-to-SQL Parsing by Self-Play Fine-Tuning](https://mlsanity.com/paper/2509.03937/) (2025)

## Abstract (arXiv)

Training a competitive Text-to-SQL agent usually depends on human-annotated natural-language/SQL pairs, which are expensive, domain-specific, and a bottleneck for scaling to new databases. We show it is possible to train a competitive solver with zero annotated pairs. We introduce SQL-Zero, a proposer-solver self-play in which a challenger and a solver start from the same base LLM and the only ground truth is execution against the database itself. The challenger generates SQL pairs calibrated to the solver's current difficulty (targeting "hard but solvable"), and both roles are updated with GRPO in alternating turns, with a template-level repetition penalty on the challenger to prevent diversity collapse. Training on BIRD databases with no labels, self-play improves over the zero-shot base on BIRD dev by 6.6 points at 3B and 7.3 points at 7B. It also scores higher than a matched control trained under the same recipe on human BIRD gold over the same databases, although an exact paired test does not resolve that margin. Transfer depends on scale: at 3B every iteration outperforms the base on unseen Spider databases and under lexical perturbation (Spider-Syn), where it also degrades less than the matched BIRD-gold control, whereas at 7B only the first iteration preserves transfer.