
Boltz Raises $110 Million as Weekly Model Use Grows Roughly 50-Fold
Boltz raised $110 million in a Series A as Andreessen Horowitz reported about 50-fold token growth after model launches; Pfizer, GSK, Takeda and Bristol Myers Squibb deploy its AI drug-design models.
- By
- Tom Whitfield
- Filed
- Channel
- Startups & Funding
- Read
- 3 min read
Boltz has raised $110 million in a Series A, less than one year after emerging from stealth with $28 million. Andreessen Horowitz co-led the seed round and invested again in the startup, which reported roughly 50-fold growth in weekly model use over four months.
Andreessen Horowitz said weekly consumption of Boltz's "molecular tokens" rose roughly 50-fold in the four months after Boltz launched BoltzMol-1 and BoltzProt-1. The investor also said Pfizer, GSK, Takeda and Bristol Myers Squibb deploy the commercial models.
The usage figure gives the round a measurable operating signal, but not a complete financial picture. The announcement provides no weekly token count, starting baseline, revenue, token price or profitability figure.
A percentage increase does not establish commercial scale without the underlying volume and pricing. The 50-fold figure therefore shows faster use, not the size of the business.
How does Boltz turn model use into revenue?
Boltz charges for model use, measured in units it calls molecular tokens. Scientists keep ownership of what they create, while the company can train exclusive model versions on a partner's proprietary data.
Under this model, more model consumption creates more billable token usage. That links charges to customer activity, although the announcement gives no rate card or revenue forecast.
Customers access the models through a web interface called Boltz Lab or an application programming interface. The API also works with coding agents such as Claude Code and Codex.
The practical result is simple: Boltz charges for use while customers keep ownership of their creations.
Which drugmakers use the commercial models?
Andreessen Horowitz names Pfizer, GSK, Takeda and Bristol Myers Squibb as commercial users of Boltz's models. That statement establishes named deployments on the investor's account, but it does not describe the size or value of each contract.
Boltz's website separately lists an agreement with Bristol Myers Squibb to deploy its biomolecular models for drug discovery. The listing provides no contract value, deployment scale or timetable.
The commercial footprint is concrete but limited in detail: four named drugmakers and one separately listed agreement.
Where did Boltz come from?
Boltz began as a research project at MIT. Gabriele Corso, Jeremy Wohlwend and Saro Passaro founded the startup.
Andreessen Horowitz said the open-source models had reached "every top 20 pharmaceutical company" by the time of the seed round. The claim describes distribution of the open models, not paid deployment across all 20 companies.
Boltz's commercial layer monetizes that activity through molecular tokens. It also offers exclusive versions trained on partner data, giving customers a company-specific deployment option.
The open models supplied broad reach, while token pricing created a way to monetize active use.
How large is the financing compared with the week's other deal?
The $110 million Series A came less than one year after Boltz launched from stealth with $28 million. Andreessen Horowitz co-led the seed round and invested again.
Boltz's round came in the same week that AI drug developer Isomorphic Labs raised $5 billion. The two transactions place a $110 million specialist raise beside a $5 billion round in the same field, but funding size does not establish equal technical or commercial progress.
Boltz has substantial venture backing, though not on the scale of the week's largest AI drug-design financing.
What must the usage growth prove?
Customer consumption is the clearest operating metric in the report. Yet the available figures do not show how many tokens customers used before the four-month period, how many organizations pay, or what Boltz earns.
The 50-fold growth can reflect a broader user base, deeper use by existing users, or both. The available information does not separate those mechanisms.
To turn the usage signal into meaningful commercial scale, Boltz must show how rising token demand translates into revenue. Its ability to train exclusive versions on proprietary partner data may support deeper adoption, although no adoption forecast accompanies the funding announcement.
The next commercial test is whether Boltz converts that 50-fold usage rise into broader paid deployments, even as the $5 billion Isomorphic Labs raise sets a much larger capital benchmark in AI drug design.
Source: mlq.ai
More from Tom Whitfield
Show full bio
Staff writer covering companies, deals and strategy in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.
1,090 articles
Related articles
model-ml-raises-250-million-at-1-5-billion-valuation
Model ML Raises $250 Million at $1.5 Billion Valuation
ramon-space-raises-124m-series-b-micron-takes-strategic-stake
Ramon.Space raises $124M Series B; Micron takes strategic stake
knowles-rebases-8-10-growth-target-from-688-million
Knowles Rebases 8-10% Growth Target to 2029 From $688 Million
tsmc-s-46-7-billion-quarter-leaves-one-question-for-investors-0facfc07
TSMC's $46.7 Billion Quarter Leaves One Question for Investors



