TAT-QA
Large-scale financial QA benchmark integrating tabular and textual data.
- Status
- UnmaintainedNo commits in 625 days.
- Last checked
- Aging · Sep 4, 2026
- License
- Open source
- Pricing
- Free tier · from $0
What it is
TAT-QA is a comprehensive question-answering benchmark focused on real-world financial datasets, combining both tabular and textual information to enhance reasoning capabilities in AI models.
TAT-QA uniquely integrates both tabular and textual financial data to provide a robust benchmark for developing and testing AI models in the financial domain.
When it fits — and when it doesn’t
A good fit for
Researchers focusing on financial data reasoning and QA tasks.
Developers building AI applications that require handling of both tabular and textual financial information.
Look elsewhere if
Projects requiring real-time streaming or low-latency responses, as TAT-QA is designed for benchmarking purposes.
General-purpose question-answering systems not focused on finance.
Known weaknesses
Each one links to the public issue, changelog or docs page it came from, where we have one.
- 01
Limited documentation and examples for complex use cases
Official documentation lacks detailed guides on advanced financial reasoning tasks
What it does well
Combines tabular and textual data for enhanced reasoning.
Focused on real-world financial datasets.
Open-source, allowing customization and integration.
Pricing
- Free tier
- YesOpen source — free to use
- Paid from
- $0
- Model
- Flat rate
- Enterprise plan
- No
Pricing not independently confirmed — check the vendor’s page.
Performance benchmarks
How Fast Is It?
Set up TAT-QA
Setup steps, common gotchas and what to check after 30 days.