Isomorphic Labs, the drug-discovery spinout of Google's DeepMind, is in talks to raise roughly $5 billion, according to Bloomberg, in a round that would value the company at about $50 billion. Other reports put the valuation in a $40 billion to $50 billion range. The round has not closed and the terms could change.
The jump is the story. Isomorphic raised $2.1 billion only five months ago in a round led by Thrive Capital. Needing capital of this size again that quickly says less about the market's appetite than about the compute and laboratory cost of a programme that wants to design medicines before it tests them.
The business underneath is IsoDDE, a platform descended from the AlphaFold work that won DeepMind a share of the 2024 Nobel Prize in Chemistry. Isomorphic does not sell pills; it sells partnerships. Research collaborations with Eli Lilly, Novartis and Johnson & Johnson use the platform to surface candidates, which the partners then take into clinical development. That is why valuations of this kind are so hard to read. A pharmaceutical company is judged on approved medicines; a platform is judged on the expectation that it structurally changes the hit rate and the duration of early discovery. For that expectation there is no public yardstick yet.
So the bet is not on a molecule. It is on the claim that predicting structure and binding will eventually compress the most expensive phase of drug development. If that link breaks in a clinical readout, the platform premium unwinds faster than it was assembled. Until then, a $50 billion valuation with no approved drug is a precedent for how the market prices scientific infrastructure, and a striking counterweight to the companies that spent the same week cutting their own internal AI budgets because they could not demonstrate what their tokens bought.
