Release notes · July 2026
PEP 2.0
PEP is the world's first AI model built for peptide research. 2.0 is the largest update since launch: the full compound library is now covered at depth, and two new capabilities make the research yours.
The full library, covered
PEP 2.0 indexes all 54 compounds in the ChatPEP library at depth. Every answer draws on the primary literature for the compound you're asking about, not a general summary of it. Ask about a mainstream compound or an obscure one and the model works from the same standard: the studies themselves, cited so you can open them.
Memory
PEP now carries your protocol across conversations. Tell it once what you're running and when you started, and every later question is answered in that context. Ask about recovery timelines in week 4 and it already knows you're in week 4.
Pulse
Research that comes to you. Pulse watches where you are in a course and surfaces what the literature says at each stage: what to expect, what's worth watching, and when it's time to reassess.
One question, before and after
Here's the difference in practice. Asked what the research says about GHK-Cu for skin, PEP 1 returned a careful general answer. PEP 2.0 returns the evidence itself: the individual studies, how they were designed, where they agree, and where the data is thin, each with a citation you can open.
What's next
Releases will keep coming: new compounds in the knowledge base, sharper retrieval, and new capabilities as the model grows.