Hebbia tests GPT-6 Astra's performance in financial research
Understand this story
Hebbia published results from testing a pre-release version of OpenAI's GPT-6 Astra in financial research scenarios: on citing the required source types, GPT-6 Astra was 13.6 percentage points higher than the best comparison model, and 99.5% of cited excerpts were relevant to the question; when generating presentations, it met 97.9% of applicable requirements, 14.4 percentage points ahead of the best comparison model, and ranked first in both briefing adherence and slide quality. The tests were conducted from August to September 2026, with each model using the same financial research cases and scoring criteria.
AI overview based on source reports · Updated 53 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- OpenAI NewsHebbia sees improvements in financial research with GPT-6 Astra
Hebbia tested OpenAI's GPT-6 Astra on two parts of its Matrix financial research workflow: selecting source documents to support answers and building PowerPoint presentations that follow a banker's brief. In Hebbia's tests, GPT-6 Astra scored 13.6 percentage points above the best alternative on citing the required types of sources, with 99.5% of cited excerpts relevant, and met 97.9% of applicable presentation requirements, 14.4 percentage points ahead of the best alternative. Hebbia's Kavita Anand said GPT-6 Astra is the first model the company has seen rank first on both presentation craft and adherence to guidelines.
Activity for this story
Not enough continuous observations to show a trend yet.