Sparrow
Structured data extraction, instruction calling and agentic workflows with ML and Vision LLM.
0 members list Sparrow, unchanged over the last 12 weeks, data as of 11 November 2026
Members who list Sparrow in their Stack, by week.
Members who list Sparrow
Be the first to add Sparrow
Nobody on Stackness runs Sparrow yet. Add it to your stack and you're the first member running it.
Homebrew analytics installs: 57 installs (2026) to 67 installs (2026)
Imported from Homebrew analytics, last reading Oct 2026
About Sparrow
Sparrow is a Python tool for extracting structured data from documents using vision and language models. Developers use it to process documents like invoices, receipts, statements and tables by uploading a file, defining a JSON schema describing the fields to extract, and receiving back structured data matching that schema. The tool supports multi-page PDFs and images, offers options like table-only extraction and schema validation, and can work with different vision model variants. It includes an agent feature that chains extraction with text-based reasoning for classification and field validation.
Sparrow is developed by katanaml and distributed under the GPL 3.0 license. The tool operates without cloud dependencies and does not store uploaded documents after processing, removing them immediately after inference completes.
Checked 9 November 2026.
Nobody on Stackness runs Sparrow yet.
Sign up and add it