What we know
Hebbia tested OpenAI’s GPT-6 Astra model for financial research and reported improvements in source coverage, citation relevance, and presentation quality during prerelease evaluations. These findings come from Hebbia’s internal assessments shared in an OpenAI customer story. Hebbia experienced better source coverage, more relevant citations, and enhanced presentation quality using GPT-6 Astra in financial research.
Why it matters
Hebbia’s use of GPT-6 Astra highlights potential advances in AI-assisted financial research, suggesting that newer models may provide more comprehensive and relevant information with clearer presentation. This could improve how financial analysts gather and interpret data. The results come from prerelease testing and are presented by OpenAI, the model’s developer.
What is still unknown
No independent tests have confirmed the accuracy or effectiveness of GPT-6 Astra in financial research. Details about supported languages, broader availability, and official release timing remain undisclosed.

