Terminal X matches senior investor judgment 88.6% with GPT-6.1
Summary
Terminal X's Eval Agent scores outputs against 54 criteria, and with GPT-6.1 Sol it catches 25% more outputs that stray from a fund's mandate.
Why this matters
Official source · Covers GPT and Agent
Source: OpenAI News · openai.com