Pathos AI, which uses AI to develop oncology drugs, raised a $237M Series E at a $3B post-money valuation, up from a $1.6B valuation in May 2025
THE SO WHAT
Pathos AI nearly doubling valuation to $3B in ~18 months on a $237M Series E is a clear vote that AI-native drug discovery—especially in oncology—is moving from story to pipeline. If you’re a traditional biotech, expect investors and partners to benchmark your discovery engine against AI-first peers on cycle time and hit quality, not just clinical endpoints.
READ THE SOURCE
MORE FROM THE WIRE
Startups & VentureArmenia’s Largest Lender Ardshinbank Is Said to Weigh London IPO
A potential London listing for Armenia’s largest lender is another data point that regional financials still see value in established Western exchanges for price discovery and governance signaling. If you’re a mid-market bank in an emerging region, the bar for staying private is rising as peers tap global equity to fund tech and risk upgrades.
Startups & VentureOpenAI Targets $1.25 Trillion Valuation in New Funding Round
A $30B raise at a targeted $1.25T valuation would turn OpenAI into a macro asset class, not just a vendor—its roadmap and pricing will directly influence indices and capital allocation. For operators, that concentration means vendor risk and platform dependency are now board-level topics, not just procurement details.
Startups & VentureJio’s Record IPO Draws Strong Demand, Defying India Market Rout
Jio’s $3.1B IPO drawing strong institutional demand despite a broader India selloff underlines how investors are still willing to pay for scaled digital infrastructure plays. If you’re building in India’s consumer or infra stack, this is a green light that public markets will reward real network effects and cashflow, even through volatility.
Startups & VentureDeepSeek reportedly raises $15bn from Tencent and CATL ahead of IPO
$15B into a single Chinese AI lab means national-scale balance sheets are now underwriting model development as strategic infrastructure. Anyone competing on foundation models without similar capital access should assume they’re playing a different game — and focus on domain, distribution, or data moats instead.