Mento benchmark: 14 AI models research startups well, but all 13 finishers rank growth worse than a headcount rule

Claude Opus 5.5 led 960 research tasks with 97.2 of 100, but every model supplied values for 9–33% of deliberately removed facts, and ranking correlations with realized growth ran from −0.16 to −0.003 against +0.06 for the heuristic.

Read at Mento.vc: Startup Agent Benchmark

More from Snape Report

OpenAI cuts its funding round target by $10 billion to about $20 billion at a $1.25 trillion valuation, FT reports
Nasdaq slides 2.9% as Intel has its worst day since the 1970s and chip stocks fall
Intel shares fall 29%, their worst day since the 1970s, cutting about $12 billion from the US government's stake
S&P revises Meta's outlook to negative over AI capital spending and negative free cash flow
Google's Suncatcher satellite returns the first AI response from a data-centre TPU in orbit
Mistral releases open weights for Mistral Large 4, its 1-trillion-parameter model
Apple says Apple Intelligence features will reach China with iOS 27.3, using Alibaba's Qwen in Siri and Writing Tools
Google, Microsoft, OpenAI and Anthropic back a shared robots.txt signal for opting out of AI training

Front page →