Terminal X Matches Senior Investor Judgment 88.6% With GPT-6.1, Cutting Query Costs 30%
Updated
Updated · OpenAI · Oct 20
Terminal X Matches Senior Investor Judgment 88.6% With GPT-6.1, Cutting Query Costs 30%
2 articles · Updated · OpenAI · Oct 20
Summary
Across 500 scored judgments, Terminal X said its Eval Agent powered by GPT-6.1 Sol agreed with senior human reviewers 88.6% of the time, 220 basis points above the next-best model.
A 54-criterion review workflow drove that result by checking each agent output against a fund’s mandate, sources, recent deals and house view, then feeding every failure back into the system as new Memory.
GPT-6.1 Sol also caught about 25% more recommendations that drifted from a fund’s strategy and ran the same long-context evaluation workflows at roughly 30% lower cost per query.
Terminal X uses forward-deployed engineers to encode firms’ unwritten investment processes into on-premise or private-cloud agents, with one private equity customer reporting $2.4 million in annualized savings.
The company said it has built 1,700 firm Memories and run more than 1 million custom workflows, and now aims to capture judgment at the individual investor level.