Astra model scores 93.2% on ParseBench for document parsing
Astra is one of the best models we've tested at parsing complex documents with tables, at 93.2% on P...
Astra's strong performance on document parsing with tables makes it a good option for finance-related document analysis, though its cost may be a factor to consider compared to other models like Fable 5.1.
Astra is one of the best models at parsing complex documents with tables, scoring 93.2% on ParseBench. It is well suited for ad-hoc analysis over table-heavy documents, such as in finance. However, it is not suitable for document-heavy OCR workloads due to its cost, at 10c per page. Fable 5.1 still performs better at charts. The model achieves 97.2% on short documents and 90.6% on medium documents on the benchmark, but only 31.7% on long documents. Its native OCR capabilities are not significantly better than other models like Fable 5.1 or gpt-5.6-sol, though it does very well on table understanding.
Astra is one of the best models we've tested at parsing complex documents with tables, at 93.2% on P...
Astra is one of the best models we've tested at parsing complex documents with tables, at 93.2% on ParseBench. So it seems well suited for ad-hoc analysis over table-heavy documents e.g. in finance. To be fair, it's also at 10c per page so not suitable for document-heavy OCR workloads. Fable 5.1 still performs a better at charts. Check out the full set of results on ParseBench: parsebench.ai Jerry Liu @jerryjliu0 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 extractbench.ai c3 🔗 View Quoted Tweet 💬 3 🔄 0 ❤️ 5 👀 754 📊 3 ⚡
