GPT-6 Astra hits 97.2% on short-document extraction, a new ExtractBench SOTA
ExtractBench found Astra best for short and medium document extraction at 97.2% and 90.6%, but long documents reach only 31.7% and cost averages about 11 cents per page.
We benchmarked GPT-6 Astra on hard document parsing and extraction tasks. It is the best frontier model for one-shotting document extraction over short and medium-sized documents: 📊 97.2% over short documents, achieving a new SOTA on our benchmark 📊 90.6% over medium documents But a few caveats: * it's pricy! The average price per page is 11c which is 10x our cost-effective document extraction solution * one-shot performance over long documents is inherently tricky, it only achieves 31.7% * the native OCR capabilities of Astra isn't that much better than other frontier models (e.g. Fable 5.1, gpt-5.6-sol). It does very well on table understanding, but isn't great on charts, layout, and semantic formatting. Full results for document extraction on ExtractBench: https://www.extractbench.ai/