DeepTable
deeptable.com
- Linux
- API
What DeepTable is
DeepTable is a cloud service that converts human-formatted Excel spreadsheets into relational, SQL-ready tables. Its LLM-powered engine interprets spreadsheet structure, relationships, merged cells, complex layouts, and visual formatting. The API accepts Excel .xlsx files up to 100MB and can return SQLite, Parquet, or CSV output. Every output value includes a reference to its source cell, allowing data to be traced back to the original sheet. Listed uses include RAG pipelines, AI agents processing business spreadsheets, SaaS spreadsheet ingestion, and enterprise internal tools. Developers can use TypeScript, Python, or Go libraries, plus a command-line tool; API and Linux are listed platforms. DeepTable says uploaded spreadsheets and processing results are not used to train or fine-tune AI models. Users choose retention periods and can delete files and results. Data is encrypted in transit and at rest. The service is in private beta, with pricing on request.
Who it suits
DeepTable suits AI engineering teams turning business spreadsheets into structured data for applications and internal tools. It is also aimed at teams needing source-cell traceability and developer libraries.
Good to know, and what to watch
- Converts complex spreadsheet layouts
- Supports SQLite, Parquet, and CSV
- Provides source-cell traceability
- Offers TypeScript, Python, and Go libraries
- Files can be deleted anytime
- Accepts Excel .xlsx input only
- Files are limited to 100MB
- Service is in private beta
- Pricing is available on request
The verdict
DeepTable focuses on a specific conversion problem: making formatted spreadsheets usable in SQL-oriented workflows. Its traceability, multiple outputs, and developer tooling are useful for pipeline builders, while private-beta availability remains a consideration.
What the maker publishes
- AI provider retention
- The privacy policy says AI providers are restricted from training on user data and may retain it for up to 30 days for abuse monitoring and legal compliance. source
- API input format
- The API reference says uploaded files must be Excel .xlsx files. source
- Cell traceability
- The site says every output value includes a reference to its source cell. source
- Company
- The legal notice identifies Totoy GmbH, based in Vienna, Austria, as the provider of DeepTable. source
- Company
- DeepTable is a service provided by Totoy GmbH, whose legal notice lists its address in Vienna, Austria. source
- Data retention
- The privacy policy says uploaded spreadsheets and processing results are kept for as long as users choose and can be deleted at any time. source
- Data retention
- Uploaded spreadsheets and processing results are kept while the user chooses to keep them and can be deleted at any time. source
- Developer tools
- The maker provides Python, TypeScript, and Go libraries, plus a CLI. source
- Headquarters
- Vienna, Austria source
- Input limit
- The site says its API accepts Excel files up to 100MB. source
- Input limit
- The API accepts Excel (.xlsx) files up to 100MB. source
- Output formats
- The site lists SQLite, Parquet, and CSV as structured output formats. source
- Output formats
- Structured data can be downloaded as SQLite, Parquet, or CSV. source
- Parsing
- DeepTable says its LLM-powered engine handles spreadsheet structure, relationships, merged cells, complex layouts, and visual formatting. source
- Parsing
- Its LLM-powered engine analyzes spreadsheet structure, relationships, and meaning, including merged cells, complex layouts, and visual formatting. source
- Privacy and AI training
- The privacy policy says DeepTable does not use uploaded spreadsheets, processing results, or other user data to train or fine-tune AI models. source
- Purpose
- DeepTable converts messy, human-formatted spreadsheets into clean, relational SQL-ready tables. source
- SDKs and CLI
- The API reference lists TypeScript, Python, and Go libraries plus a CLI tool. source
- Security
- The privacy policy says data is protected with industry-standard encryption in transit and at rest. source
- Security
- The privacy policy says data is encrypted in transit and at rest, and uploaded user data is not used to train or fine-tune AI models. source
- Support and availability
- The maker says DeepTable is in private beta and invites AI engineering teams to book a demo; a support email is listed as [email protected]. source
- Supported input
- The API documentation says uploaded files must be Excel (.xlsx) format. source
- Traceability
- The product says each output value includes a reference to its source cell. source
- Use cases
- The site describes use cases including RAG pipelines, AI agents that process business spreadsheets, SaaS spreadsheet ingestion, and enterprise internal tools. source
Checked against the maker's pages on 9 Oct 2026.