tablegpt/TableGPT-R1

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Dec 23, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

TableGPT-R1, developed by Zhejiang University & Institute of Computing Innovation, Zhejiang University, is a specialized large language model built on the Qwen3-8B architecture with a 128K token context window. It is optimized for complex tabular reasoning and data analysis, utilizing a Reinforcement Learning (RL) framework for autonomous agentic reasoning and robust code execution. This model excels at multi-step logic and environment interaction, particularly with table-path inputs and a built-in code interpreter, and shows strong performance in NL2SQL and holistic table evaluation benchmarks.

Loading preview...

TableGPT-R1: Advanced Tabular Reasoning with Reinforcement Learning

TableGPT-R1 is a specialized large language model from Zhejiang University, designed for complex tabular data analysis and reasoning. Built upon the Qwen3-8B transformer architecture, it features an extended 128K token context window and a specialized tokenizer for efficient handling of tabular data and code syntax. Unlike traditional models, TableGPT-R1 leverages a systematic Reinforcement Learning (RL) framework, enabling it to function as an autonomous agent capable of multi-step logic, robust Python/SQL code execution, and iterative refinement based on environment feedback.

Key Capabilities

  • Autonomous Agentic Reasoning: Generates visible reasoning chains within <think> tags, plans data manipulations, and refines strategies using a Code Interpreter.
  • Unified Reward System: Employs a hybrid reward mechanism combining rule-based verification with a Criteria-Injected Reward Model for accuracy and interpretability.
  • GRPO++ Framework: Optimizes decision-making across diverse table structures while maintaining general-purpose reasoning.
  • Table-Path Inputs: Autonomously loads and retrieves information from files using a built-in code interpreter.
  • Agentic Loop Integration: Supports a seamless "Think-Act-Observe" cycle, treating environment feedback as a first-class input for real-time error correction.

Good for

  • Complex Tabular Data Analysis: Excels at multi-table joins, hierarchical reasoning, and data processing.
  • Natural Language to SQL/Code: Demonstrates superior generalization, with significant performance increases on Spider and BIRD benchmarks compared to TableGPT2-7B.
  • Autonomous Data Science Workflows: Ideal for tasks requiring iterative code execution, error correction, and logical deduction in data environments.
  • Chinese Language Tabular Queries: Strong emphasis on Chinese corpora, though other languages may have limited support.

TableGPT-R1 shows substantial advancements, outperforming Qwen3-8B and even GPT-4o in specific RealHitBench tasks, particularly in Chart Generation. More details are available in the research paper.