Iambic Therapeutics has no approved drug. Its lead molecule is in Phase 1. Seven years of work, roughly $462 million of private capital, and the company's most advanced asset has cleared only early safety testing.
On Wednesday, it raised $150 million anyway.
The San Diego company sold 9,375,000 shares at $16.00 in a Nasdaq initial public offering, pricing at the midpoint of its marketed range and valuing it near $758 million. Trading opened Thursday under the ticker IAM. Existing investors ARK Investment Management and Duquesne Family Office had indicated they would take up to $60 million of the offering.
The gap between money and proof defines the AI-drug boom. Iambic just widened it in public.
The bet is that AI-designed molecules close the distance between lab and clinic faster than rivals can. Nothing in the filings yet shows that distance shrinking.
Where the $150 million goes
Iambic's prospectus is unusually specific about allocation. About $75 million will fund midstage development of IAM1363, an oral, brain-penetrant HER2 inhibitor now in Phase 1 for HER2-positive breast cancer and non-small cell lung cancer.
A further $15 million each goes to IAM217, a KIF18A inhibitor whose investigational new drug application was filed on Oct. 5, and to IAM-C1, a CDK2/4 candidate for HR+/HER2- breast cancer. The final $20 million is earmarked for the AI platform itself.
Three clinical programs. None approved. All early. The offering is less a wager on one molecule than on the engine that keeps producing them.
The engine, and what it has sold
That platform runs on Enchant v3 — a multimodal transformer model trained, the company says, on more than 6,000 molecular properties — alongside NeuralPLexer for structure prediction. Iambic markets the pairing as "molecular superintelligence."
Large pharma has bought the pitch. The company holds discovery partnerships with AbbVie, Bayer, Takeda, Lundbeck, Revolution Medicines and Jazz Pharmaceuticals, the AbbVie agreement signed as recently as September. Those deals bring near-term revenue and external validation while the internal pipeline matures.
The more instructive precedent is the last AI-bio cohort that reached the public markets before its pipelines matured. Investors learned then that a model can narrow the search for molecules. It cannot compress a Phase 3 trial.
So why do investors keep paying for that gap?
Because the prizes are enormous and the downside is partly shared. A pharma partner that funds the search absorbs much of the cost, and Iambic keeps the option on whatever comes out the far end.
What to watch
Two signals will decide whether this IPO reads as a milestone or a warning. The first is Phase 1 data for IAM1363 — brain penetration and objective response in a HER2 population that is not short of existing treatments. The second is whether the platform partnerships convert into milestone payments, the only outside proof that AI-designed molecules are worth paying for at scale.
Iambic has now raised $462 million privately and $150 million in the public market. The slow, expensive part of drug discovery — the years between a promising molecule and a proven one — is precisely where the money is meant to go. It is also the part no model has yet been shown to shorten.