How to Use Mistral document text and table extraction in Claude Code
Extract text, embedded images, and reconstructed tables from PDFs and images into Markdown. Use it to prepare reports, scans, and technical papers for search or analysis while retaining document structure.
Prompts to try with Mistral document text and table extraction in Claude Code
Copy an example into Claude Code, or ask in your own words.
Use Mistral document text and table extraction on https://arxiv.org/pdf/1706.03762 and save the page-level Markdown under docs/papers/attention/.
Use Mistral document text and table extraction on https://arxiv.org/pdf/1810.04805 and build a CSV of the extracted tables, preserving their full column headings.
Use Mistral document text and table extraction on https://arxiv.org/pdf/1512.03385 and add its extracted text to this project's search index with page references.
Use Mistral document text and table extraction on https://arxiv.org/pdf/1706.03762 and create a project fixture containing the extracted images and their page associations.
Use Mistral document text and table extraction on https://arxiv.org/pdf/1810.04805 and write a JSON document record containing its extracted title, section headings, and table captions.
How to set up Mistral document text and table extraction in Claude Code
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
This adds Spiral to the current project. Add --scope user to use it in every project.
claude mcp add --transport http spiral https://api.spiral.ai/mcp
- Sign in
Start claude, type /mcp, choose spiral and Authenticate. Sign in to Spiral in the browser and click Authorize. No API key.
Use the Spiral skill
- Send Claude Code this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run Mistral document text and table extraction
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "mistral", "model": "mistral-ocr-2512", "input": { "document": { "document_url": "<document_url>" } } }'
Then paste one of the prompts above: Claude Code finds Mistral document text and table extraction on Spiral, shows the price and runs it.
Mistral document text and table extraction providers and prices
Spiral runs Mistral document text and table extraction through Mistral. Open a card for today's price, with every fee included.
Settings you can ask for
Mistral document text and table extraction works from your document annotation prompt. Ask Claude Code for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Image limit | a whole number | Max images to extract | |
| Table format | markdown or html | ||
| Extract footer | on or off | off | Extract the page footer into the response's |
| Extract header | on or off | off | Extract the page header into the response's |
| Image min size | a whole number | Minimum height and width of image to extract | |
| Include blocks | on or off | on | Return paragraph-level bounding boxes for all content blocks in the response |
| Include image base64 | on or off | Include image URLs in response | |
| Confidence scores granularity | word, page, or block | Granularity for confidence scores: 'page' (aggregate only), 'word' (per-word scores), or 'block' (per-block scores). Defaults to None (no confidence scores) to keep response payload small. |
Every API field of Mistral OCR 3 (2512) is on its tool page: open its card above.
What can you do with Mistral document text and table extraction in Claude Code?
Build a searchable collection of technical PDFs with page-level Markdown and extracted tables. A coding agent can put Mistral OCR results into the project's document index and preserve table structure for later analysis.
Make public reports searchable
Convert annual reports and agency publications into page-level Markdown for a searchable reference collection.
Recover tables for analysis
Extract tables from public research papers and business reports, retaining merged headers before converting the results into analysis-ready rows.
Build a technical reference assistant
Turn equipment manuals and technical PDFs into text and embedded images that an assistant can use to answer document questions.
Compare report editions
Extract successive editions of a public report, then compare sections and tables to identify changes in the published material.
Create structured document records
Use document annotations to collect selected headings, report identifiers, and other document details into a consistent JSON structure.
Prepare research reading notes
Extract a paper's text, tables, and figures before producing notes that connect claims to their source pages.
More on Mistral document text and table extraction and Claude Code
Overview
Mistral document text and table extraction in other agents
Questions about Mistral document text and table extraction in Claude Code
Can Claude Code use Mistral document text and table extraction?
Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run Mistral document text and table extraction like any built-in tool.
How much does Mistral document text and table extraction cost in Claude Code?
$0.0026 per page ($0.0039 with document or bounding-box annotations). You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude Code plan is separate.
Do I need a Mistral account to use Mistral document text and table extraction in Claude Code?
No. Spiral runs Mistral document text and table extraction on its own Mistral account. You sign in to Spiral once and never paste a key.
How long does it take to set up Mistral document text and table extraction in Claude Code?
About two minutes: add the hosted server, then sign in.
Does Claude Code wait while Mistral document text and table extraction runs?
Mistral document text and table extraction runs in the background. Claude Code starts the run, checks on it and gives you the json when it's ready.
Which document formats can Mistral OCR process?
Mistral lists PDF and seven image formats: PNG, JPG, JPEG, TIFF, BMP, GIF, and WEBP. The extracted content comes from the document you submit.
How many pages can one OCR call process?
Mistral's document chunking cookbook gives approximately 1,000 pages as the typical maximum per inference call. It recommends splitting documents that exceed the limits and combining the extracted results afterward.
Can Mistral OCR preserve complex tables?
OCR 3 reconstructs table headers, merged cells, multi-row blocks, and column hierarchies. Its HTML table output can represent structures that a simple Markdown table cannot.
Does OCR 3 return paragraph bounding boxes?
No. Mistral documents paragraph-level block extraction as an OCR 4 or newer feature, so OCR 3 should not be used for paragraph-coordinate workflows.
Does the reported 74% result mean 74% text accuracy?
No. Mistral reports a 74% overall win rate against OCR 2 across forms, scanned documents, complex tables, and handwriting. That measures comparative wins, not the percentage of characters transcribed correctly.
Start using Mistral document text and table extraction in Claude Code
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Claude Code:
set up https://spiral.ai/SKILL.md


