---
papers_schema: "1.0"
title: "AI-readable financial models (Excel)"
canonical: "https://thepapers.link/use-cases/financial-models"
status: "owner_published"
---

# Give AI your financial model without losing the formulas

Publish an Excel financial model as a verified AI Version: values, number formats, formulas, dependencies and assumptions preserved, at a permanent link.

## The problem

A model exported to PDF loses everything that makes it a model: which cells are inputs, which are formulas, and what depends on what.

Pasted into a chat, a spreadsheet becomes a grid of numbers with no units, periods or assumptions, and AI fills the gaps by guessing.

## How The Papers handles it

1. Upload the .xlsx. It is read natively — never converted to PDF first.
2. Sheets, values with their number formats, formulas, dependencies, named ranges and assumptions go into the canonical model.
3. Every published value is checked against its source cell; formulas must be preserved and their dependencies resolved.
4. Agents read the JSON (metrics with period, currency and classification, each with its source cell) or the Markdown.

## What is checked against the source

- cell value exact match
- formula preserved
- dependencies resolved
- period and currency on each metric
- inputs vs calculated values

## Where to put the link

The cover sheet of the model, the email you send it with, or your data room.

---

The Papers is not a PDF-to-Markdown converter. Conversion is the first step. Before an AI Version is published, its figures, currencies, dates, periods, qualifiers and formulas are checked against the original document, each value keeps a reference to its source page or cell, and anything the source does not settle is put to the document owner instead of guessed.

[What The Papers is](https://thepapers.link/product.md) · [Pricing](https://thepapers.link/#pricing) · [Start free](https://thepapers.link/)