OpenAI’s API pricing page now lists gpt-6-astra-law under “Legal model configurations” at $10.00 per million input tokens, $50.00 per million output tokens and $1.00 for cached input. That is the same price the page gives GPT-6 Astra itself.
The page adds that “GPT-6 Astra for Law promotional pricing is available at least through December 31, 2026”. By our arithmetic the legal configuration costs 1.0x GPT-6 Astra, while the page prices OpenAI’s life-sciences model at 1.5x GPT-6 Astra’s price, and its cyber model at 3.1x the input price of gpt-5.6-sol, listed beside it.
Business Pill · THE MODEL AND THE HARNESS
A short Business Engineer explainer on the difference between an AI model and the harness of tools and workflow around it. It teaches the general idea only and does not discuss Astra for Law or its price.
The key insight: OpenAI’s pricing page lists gpt-6-astra-law at $10.00 input and $50.00 output per million tokens, the same as GPT-6 Astra, on promotional terms at least through 31 December 2026; by our arithmetic that is 1.0x, against 1.5x for its life-sciences model.
TestingCatalog spotted the listing on 31 October, writing on X that gpt-6-astra-law “is now listed on the API pricing page”. We read the page the same morning.
In the flagship table, gpt-6-astra costs $10.00 input, $50.00 output and $1.00 cached input per million tokens at standard speed. The legal table lists gpt-6-astra-law at the same three prices.
Two other specialist tables sit on the same page. The life-sciences model gpt-6-rosalind is listed at $15.00 input and $75.00 output, and the page says “Access is limited to approved internal research through the trusted-access program”. The cyber model gpt-5.6-cyber is listed at $12.50 input and $75.00 output, next to gpt-5.6-sol at $4.00 and $20.00.
Input price per million tokens, specialist model vs a comparison model (gpt-6-astra from the flagship table; gpt-5.6-sol from the cyber table), OpenAI API pricing page, read 31 Oct 2026: gpt-6-astra-law $10.00 vs gpt-6-astra $10.00 (1.0x); gpt-6-rosalind $15.00 vs gpt-6-astra $10.00 (1.5x); gpt-5.6-cyber $12.50 vs gpt-5.6-sol $4.00 (3.1x). Ratios are our arithmetic; the legal price is promotional at least through 31 Dec 2026.
What OpenAI Said on 17 September
OpenAI introduced Astra for Law on 17 September as a configuration that “combines GPTโ6 Astra with a powerful legal search index and instructions for legal analysis and writing”. It said the product would be “initially offered to selected law firms through Trusted Access in ChatGPT and Codex, and will be coming soon to the API”, appearing “in the API as gpt-6-astra-law”.
OpenAI also said “API customers including Harvey and Legora will be able to build on Astra for Law”. On Vals AI’s Legal Research Bench validation set, it reported that Astra for Law “passed the evaluationโs overall correctness check on 54.0% of questions, compared with 38.7% for GPTโ6 Astra using web search alone”.
When we covered the Astra for Law launch on 18 September, the announcement gave no price. The pricing page now gives one, 44 days later by our count.
The Business Lever: Specialist Configurations Without a Specialist Premium
The lever here is pricing. By our arithmetic, the life-sciences model gpt-6-rosalind costs 1.5x GPT-6 Astra in the flagship table, and the cyber model costs 3.1x on input against gpt-5.6-sol in its own table. The legal configuration, for now, costs the same as GPT-6 Astra.
OpenAI said on 17 September that API customers including Harvey and Legora would be able to build on Astra for Law. The page now prices that configuration at GPT-6 Astra’s rate. AI-native legal firms are also moving on the incumbents, as with Crosby’s bid for Weil Gotshal’s brand and data.
The Structural Read
As we read the pricing page, the legal configuration is listed in its own table, separate from the flagship models.
As we read OpenAI’s 17 September announcement, Astra for Law is GPT-6 Astra plus a legal search index and instructions, not a separately trained model.
As we read the promotional sentence, the page sets a floor date for the current price, not a later price.
OpenAI, Introducing Astra for Law, 17 September 2026
“API customers including Harvey and Legora will be able to build on Astra for Law, bringing this intelligence into their own products and workflows.”
Three Implications
FOR LEGAL-AI BUILDERS OpenAI said API customers including Harvey and Legora will be able to build on Astra for Law.
FOR LAW FIRMS OpenAI said Astra for Law would be initially offered to selected law firms through Trusted Access in ChatGPT and Codex.
FOR PRICING The pricing page lists the legal configuration on promotional terms at least through 31 December 2026.
The framework says “The agentic surface โ the harness โ is becoming where value pools in the entire AI stack.”
As we read it, Astra for Law fits that frame: OpenAI describes it as GPT-6 Astra plus a legal search index and instructions. The framework predates the listing and does not discuss legal models or this price.
What Is Not Established
The pricing page lists the configuration and its price. We did not find a separate OpenAI announcement saying gpt-6-astra-law is generally available on the API, or that the Trusted Access limits for law firms have changed. The page does not say what the price becomes after the promotional period.
Wayback Machine copies of the pricing page do not include the rendered price tables, so we could not confirm the date the legal line first appeared. We date it from TestingCatalog’s post and our own reading on 31 October.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
OpenAI’s pricing page lists gpt-6-astra-law at $10.00 input and $50.00 output per million tokens, the same as GPT-6 Astra, on promotional terms at least through 31 December 2026. By our arithmetic that is 1.0x GPT-6 Astra, against 1.5x GPT-6 Astra for OpenAI’s life-sciences model and 3.1x gpt-5.6-sol’s input price for its cyber model.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Prices come from OpenAI’s API pricing page as we read it on 31 October 2026; launch details come from OpenAI’s 17 September announcement. Price ratios are our arithmetic. We did not contact OpenAI.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
