Timeline

OpenAI launches Astra for Law

OpenAI reported 54.0% accuracy on Vals AI's Legal Research Bench with the configured system, against 38.7% for plain GPT-6 Astra using web search.

  • Models & capabilities
  • Minor

OpenAI launched Astra for Law, a version of GPT-6 Astra configured for legal research, built around a case-law search index assembled with the nonprofit Free Law Project. The index draws heavily on the Free Law Project’s CourtListener collection, which OpenAI said covers more than 99.9% of published US precedential case law, combined with statutes, regulations, court rules and administrative decisions across more than 230 million URLs.

The product launched alongside 26 partner plugins, including connectors from Thomson Reuters and integrations from legal-AI firms Harvey and Legora, plus legal-software vendors such as Relativity, Clio and iManage. OpenAI reported that, tested on 200 US legal research questions from a private validation set of Vals AI’s Legal Research Bench, the configured Astra for Law system passed 54.0% of questions at its highest reasoning setting, against 38.7% for plain GPT-6 Astra using web search alone — a relative improvement, though the system still failed nearly half the questions tested.

Access at launch was limited to selected law firms through a “Trusted Access” programme in ChatGPT and Codex; OpenAI said API access, under the identifier gpt-6-astra-law, would follow, and did not publish pricing or a list of participating firms. The launch covers US law only.

Read next

In the commentary

What people were saying around this time — external links, from the record's commentary rail.