# Unstructured launches Transform models that out‑perform GPT‑6, Claude and Gemini

> In tests on 224 business pages, Unstructured's new Transform models made far fewer table errors and hallucinations than leading AI models.

Oossa · 2026-10-21 · https://oossa.com/en/unstructured-launches-transform-models-that-out-perform-gpt-6-claude-and-gemini

Unstructured released upgraded Transform models for its document‑parsing service. The company tested the models on 224 real‑world business pages and found they produced up to 2.3 × fewer table errors than GPT‑6, Claude and Gemini. The tests also showed the competing models hallucinated about 50 % more content that never appeared in the source documents. Unstructured says the new models are ready for customers who need reliable data extraction from PDFs, scans and complex forms.

## The facts

- Test set: 224 real‑world business pages
- Error reduction: up to 2.3 × fewer table errors than GPT‑6, Claude and Gemini

## Why it matters

Businesses that rely on automatic data extraction can expect more accurate tables and fewer hallucinated entries, reducing manual correction work.

## Sources & references

1. [Unstructured's New Transform Models Beat GPT-6, Claude and Gemini at Parsing Tables, Forms and Scans](http://www.businesswire.com/news/home/20261021388350/en/Unstructureds-New-Transform-Models-Beat-GPT-6-Claude-and-Gemini-at-Parsing-Tables-Forms-and-Scans/?feedref=JjAwJuNHiystnCoBq_hl-WsGkFyUNMTSNrPsg57HLEqqcp-o_pnudlUwsb5apQ1S4gUE65BTfjH3-pSuqdv0gW3cb3F4oTIgUqCPafFkgu5AneicJQDNboWvodvG82E0) – Business Wire, 2026-10-21

Last updated: 2026-10-21