OpenEvidence hits $15 billion as studies question its accuracy

OpenEvidence has attracted doctor adoption organically, which Thrive Capital partner Nworah Ayogu said gives it a degree of differentiation. That adoption comes despite mixed evidence on its effectiveness, with the Mayo Clinic study positive on accuracy but limited on treatment impact, and a June Nature paper finding it underperformed general-purpose models from Google and OpenAI on clinical benchmarks.

The competitive landscape is crowded. OpenAI has made a play for healthcare with ChatGPT, and Anthropic has released healthcare tools, though it recently partnered with OpenEvidence to bring those tools to 100 low- and middle-income countries. Eric Oermann, a neurosurgeon at NYU Langone, noted a lack of research into individual AI tools and scant guidance on clinical implementation, while concerns exist that junior doctors could rely too heavily on such systems without mastering basic knowledge.

Nadler framed the end game as literature-based drug discovery, saying its economic value could dwarf what the company could charge hospitals or advertisers. That ambition, alongside the mixed accuracy findings, suggests the company's path forward may hinge on proving clinical value as it expands beyond its ad-supported core.