Spider 2.0: Evaluating Language Models on Real-World Enterprise Text-to-SQL Workflows
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:V6S53V4Xrecord.jsonopen to challenge →
read the original abstract
Real-world enterprise text-to-SQL workflows often involve complex cloud or local data across various database systems, multiple SQL queries in various dialects, and diverse operations from data transformation to analytics. We introduce Spider 2.0, an evaluation framework comprising 632 real-world text-to-SQL workflow problems derived from enterprise-level database use cases. The databases in Spider 2.0 are sourced from real data applications, often containing over 1,000 columns and stored in local or cloud database systems such as BigQuery and Snowflake. We show that solving problems in Spider 2.0 frequently requires understanding and searching through database metadata, dialect documentation, and even project-level codebases. This challenge calls for models to interact with complex SQL workflow environments, process extremely long contexts, perform intricate reasoning, and generate multiple SQL queries with diverse operations, often exceeding 100 lines, which goes far beyond traditional text-to-SQL challenges. Our evaluations indicate that based on o1-preview, our code agent framework successfully solves only 21.3% of the tasks, compared with 91.2% on Spider 1.0 and 73.0% on BIRD. Our results on Spider 2.0 show that while language models have demonstrated remarkable performance in code generation -- especially in prior text-to-SQL benchmarks -- they require significant improvement in order to achieve adequate performance for real-world enterprise usage. Progress on Spider 2.0 represents crucial steps towards developing intelligent, autonomous, code agents for real-world enterprise settings. Our code, baseline models, and data are available at https://spider2-sql.github.io
This paper has not been read by Pith yet.
Forward citations
Cited by 28 Pith papers
-
EntSQL: A Benchmark for Grounding Text-to-SQL in Long-Context Enterprise Knowledge
EntSQL is a new benchmark with 1,066 examples across five domains where top systems reach only 15.9% accuracy on English inputs when long-form enterprise documents are provided.
-
Residual Skill Optimization for Text-to-SQL Ensembles
Residual skill optimization creates complementary Text-to-SQL agents by training each new skill on prior ensemble failures, yielding accuracy gains on Spider2-Lite and transfer to other dialects and tasks.
-
LEAF-SQL: Level-wise Exploration with Adaptive Fine-graining for Text-to-SQL Skeleton Prediction
LEAF-SQL uses level-wise exploration with adaptive fine-graining and dual agents to generate diverse SQL skeletons, reaching 71.6% execution accuracy on the BIRD benchmark and outperforming prior search- and skeleton-...
-
DataClawBench: An Agent Benchmark for Exploratory Real-World Financial Data Analysis
Introduces DataClawBench benchmark for exploratory financial data analysis by agents and reports that exploration does not reliably improve task outcomes in noisy cross-domain settings.
-
DataClawBench: An Agent Benchmark for Exploratory Real-World Financial Data Analysis
DataClaw supplies a process-oriented benchmark of real-world noisy data and milestone-annotated tasks that shows seven of eight tested LLMs achieve below 50% accuracy on exploratory analysis.
-
EXPO-SQL: Execution-based Clause-level Policy Optimization for Text-to-SQL
EXPO-SQL improves Text-to-SQL by using clause-level rewards derived from execution error messages and incremental clause execution instead of uniform query-level rewards.
-
SynQL: A Controllable and Scalable Rule-Based Framework for SQL Workload Synthesis for Performance Benchmarking
SynQL synthesizes diverse, execution-ready SQL workloads by deterministically traversing foreign-key graphs to populate ASTs, yielding high topological entropy and cost-model training data with R² ≥ 0.79 on held-out sets.
-
SpotIt+: Verification-based Text-to-SQL Evaluation with Database Constraints
SpotIt+ uses verification to find realistic counterexample databases that expose discrepancies between generated and gold SQL queries missed by standard test-based evaluation on the BIRD dataset.
-
Both Ends Count! Just How Good are LLM Agents at "Text-to-Big SQL"?
New Text-to-Big SQL metrics show that LLM agents must balance accuracy with cost and speed at scale, where GPT-4o trades some accuracy for up to 12x speedup and GPT-5.2 proves more cost-effective than Gemini 3 Pro on ...
-
Spider 2.0-AIFunc: Extending Real-World Text-to-SQL to AI-Native SQL Workflows
Spider 2.0-AIFunc is a 465-instance benchmark for evaluating text-to-SQL systems on queries that incorporate Snowflake Cortex AI functions, with evaluations of ten models showing proprietary models reach 67-70% accuracy.
-
Database Context Compression for Text-to-SQL on Real-World Large Databases
DBCC applies SGCF-based offline compression and online purification to shrink database context by up to 100x while raising schema recall and execution accuracy 1.8-1.9% on Spider 2.0 and BIRD.
-
EcoTable: Cost-effective Table Integration in Data Lakes for Natural Language Queries
EcoTable is the first NL-based data integration framework that builds a join-likelihood graph, uses two-stage schema linking and Steiner tree search to find paths, then generates transformations with LLMs, reporting >...
-
SOMA-SQL: Resolving Multi-Source Ambiguity in NL-to-SQL via Synthetic Log and Execution Probing
SOMA-SQL resolves multi-source ambiguity in NL-to-SQL using synthetic query logs and ambiguity-driven execution probing, reporting 13% average execution accuracy gains over baselines on six benchmarks.
-
Cross-Vendor Sola ISPM Benchmark: Evaluating Agentic AI for Federated Identity Security Reasoning
Presents the Cross-Vendor Sola ISPM Benchmark and reports that adding relational context raises AI answer correctness by 34% and cuts exploration queries by 70% on multi-vendor identity tasks.
-
Towards Direct Evaluation of Harness Optimizers via Priority Ranking
Priority ranking offers a low-cost direct evaluation for harness optimizers that correlates with their real multi-step optimization performance, supported by the Shor dataset of 182 scenarios.
-
Hypergraph Enterprise Agentic Reasoner over Heterogeneous Business Systems
HEAR uses a stratified hypergraph ontology to orchestrate evidence-driven multi-hop reasoning over heterogeneous business systems, reaching 94.7% accuracy on supply-chain root-cause tasks with open-weight models.
-
Anatomy of a Query: W5H Dimensions and FAR Patterns for Text-to-SQL Evaluation
Text-to-SQL queries universally reduce to Filter-Aggregate-Return operations with domain-varying W5H semantic profiles, showing near-zero causal and mechanistic reasoning everywhere.
-
DataClawBench: An Agent Benchmark for Exploratory Real-World Financial Data Analysis
DataClawBench is a new benchmark for exploratory real-world financial data analysis that shows increased exploration by LLM agents does not reliably produce task-relevant progress or correct answers.
-
From Unstructured Recall to Schema-Grounded Memory: Reliable AI Memory via Iterative, Schema-Aware Extraction
Schema-aware iterative extraction turns AI memory into a verified system of record, reaching 90-97% accuracy on extraction and end-to-end memory benchmarks where retrieval baselines score 80-87%.
-
SemanticAgent: A Semantics-Aware Framework for Text-to-SQL Data Synthesis
SemanticAgent introduces a three-stage semantic analysis, synthesis, and verification process that produces higher-quality text-to-SQL training data than prior execution-only methods.
-
AV-SQL: Decomposing Complex Text-to-SQL Queries with Agentic Views
AV-SQL uses a pipeline of LLM agents to generate intermediate CTE views that decompose complex Text-to-SQL queries, reaching 70.38% execution accuracy on Spider 2.0.
-
LLMs Get Lost In Multi-Turn Conversation
LLMs drop 39% in performance during multi-turn conversations due to premature assumptions and inability to recover from early errors.
-
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
-
AgentNLQ: A General-Purpose Agent for Natural Language to SQL
A multi-agent LLM framework with schema enrichment and business rules achieves 78.1% semantic accuracy on the BIRD NL2SQL benchmark.
-
Consistency as a Testable Property: Statistical Methods to Evaluate AI Agent Reliability
A framework with U-statistics and kernel-based metrics quantifies AI agent consistency and robustness, showing trajectory metrics outperform pass@1 rates in diagnosing failures.
-
From Natural Language to PromQL: A Catalog-Driven Framework with Dynamic Temporal Resolution for Cloud-Native Observability
A catalog-driven framework translates natural language into PromQL queries with dynamic temporal resolution for cloud-native observability.
-
BADGER: Bridging Agentic and Deterministic Evaluation for Generative Enterprise Reasoning
BADGER is a new enterprise evaluation framework that adds LLM-assisted SQL component extraction and a Hybrid-EX metric validated on 150 human-annotated queries to existing text-to-SQL and agentic assessment methods.
-
ClinQueryAgent: A Conversational Agent for Population Health Management
The paper introduces ClinQueryAgent, a conversational agent that converts natural language queries into database queries for population health management while keeping patient data secure, and reports its use by 128 s...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.