Terminal X Reports 88.6% Agreement With Senior Investors
OpenAI published a customer story on October 20, 2026 about Terminal X, a startup that sells what it calls a Capital Memory operating system to investment firms. Everything below comes from that first-party page on OpenAI's own site.
Transcript
Terminal X says its agent now matches senior investor judgment. OpenAI published the numbers.
Terminal X encodes how a firm invests into agents the firm owns. When senior investors leave, the firm keeps documents but loses reasoning.
On GPT 6.1 Sol its Eval Agent grades output against fifty four criteria. It agreed with the human reviewer eighty eight point six percent.
Terminal X reports one customer saved two point four million dollars a year on due diligence. It also reports lower cost per query.
These are Terminal X's own numbers, published by OpenAI as a customer story. The model compared against is never named.
Inboxsmith builds AI receptionists for small business. Please like and subscribe for more news.
Sources
Every claim in this video comes from the top ranking coverage of this topic. The claims and where each one came from:
- Terminal X's Capital Memory operating system encodes how a firm invests into agents the firm owns, deployed in the firm's own private data room on premise or in its own cloud, with every production query auditable.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)
- Terminal X's Eval Agent checks each agent output against a 54-criterion rubric, scoring each check pass or fail.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)
- Across 500 scored judgments the Eval Agent on GPT-6.1 Sol agreed with the senior human reviewer 88.6% of the time, versus 86.4% for the unnamed next-best model, an improvement of 220 basis points.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)
- According to Terminal X, one private equity firm reported $2.4 million in annualized savings from automated due diligence and investment memo workflows.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)
- GPT-6.1 Sol ran the same workflows at about 30% lower cost per query than the next-best model and caught about 25% more recommendations that strayed from a fund's mandate.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)
- The performance figures are self-reported by Terminal X and published by OpenAI on its own site as a customer story; the comparison model is never named.(OpenAI (openai.com), first party customer story titled "Terminal X matches senior investor judgment 88.6% with GPT-6.1", October 20, 2026)