The use case: how European Patent Office deploys Mistral
The EPO uses a fine-tuned OCR model from Mistral, with a pipeline that outputs ST36 XML, to automatically transcribe complex patent documents containing formulas, tables and figures. The solution is built for on-premises deployment.
More reported results
Figures published in Mistral’s case study. They are not independently measured by usedby.
<1%
<1% character recognition error rate
5 days
Lead time from up to 5 days down to a few minutes
In the customer’s words
Standard OCR tools often fall short when confronted with the intricacies of patent documents. What mattered to us was ST36 compliance and reliability on complex elements like tables, formulas, and figures at high volumes. Mistral AI's OCR expertise directly addressed the near-zero fault tolerance our patent grant process requires.Mistral users in Government & GovTech
Companies in the same sector, with public evidence.
See all Mistral customersGo straight to the source
Based on Mistral’s published case study. We check the public evidence; the results are reported by the vendor.
mistral.ai- Source published
- Apr 24, 2026
- Last checked
- Nov 10, 2026
Open source capture What the story claims, and what we checked
What we compared with the page.
- The figure: 400,000 pagesChecked
The figure is on the page; its label is our wording.
- The publication dateChecked
Read from the page’s own metadata, never guessed.
- European Patent Office uses MistralChecked
Confirmed line.
- The result itselfNot checked
We quote it; we did not measure it.
Specifically, the system achieved a throughput of 400,000 pages transcribed per day while maintaining a character recognition error rate of less than 1%.

