Dimension Capital Closes $800M Fund III, Doubling Down on Science and Compute

Dimension Capital announced an $800 million third fund on July 21, 2026, scaling its investment focus at the intersection of scientific research and advanced computation. The New York-based firm, founded in late 2022 by Zavian Dar, Adam Goulburn, and Nan Li, has now materially expanded its capital base in a compressed timeframe.
The third fund is 60 percent larger than the firm's $500 million second fund, which was announced in December 2024 and had itself exceeded a $400 million target. Fund III arrives approximately 18 months after Fund II. Prior to launching Dimension, Dar and Goulburn were partners at Lux Capital, while Li was at Obvious Ventures.
Earlier reporting tracked the trajectory of Fund III's formation. STAT News reported on March 17, 2026, that Dimension was scouting $700 million for a new fund. Endpoints News subsequently reported on June 4, 2026, that the firm was seeking up to $750 million. The final $800 million figure surpassed both of those reported targets.
Dimension's existing portfolio already provides concrete evidence of where capital has been deployed. Dimension co-led a $30 million seed investment in Chai Discovery in 2024. Chai Discovery develops open-source AI foundation models for drug development and, as of July 2026, had raised $400 million at a $3.8 billion valuation. Dimension also backed New Limit, co-founded by Coinbase CEO Brian Armstrong, at its Series A round in January 2025. New Limit completed a Series C at a $3.1 billion valuation as of July 2026.
On the compute infrastructure side, Dimension invested in Modal Labs. The firm also became an Anthropic shareholder when Anthropic acquired Dimension portfolio company Coefficient Bio in spring 2026 for approximately $400 million. Coefficient Bio operated as a drug discovery platform. Dimension received Anthropic shares through that acquisition transaction.
The Coefficient Bio acquisition illustrates an emerging pattern: frontier AI companies acquiring computational biology assets not merely for talent, but for domain-specific scientific platforms that can be integrated into broader model development pipelines. The distance between a $30 million seed check in 2024 and a $3.8 billion valuation for Chai Discovery in mid-2026 reflects how rapidly open-source foundation models for drug development are being valued. The premium that Anthropic placed on Coefficient Bio, an approximate $400 million acquisition price, signals that AI labs are willing to pay directly for proprietary biological data and discovery infrastructure rather than building those capabilities from scratch.
For limited partners, Dimension's fund trajectory, from a $500 million Fund II to an $800 million Fund III in 18 months, reflects a broader reallocation of venture capital toward firms with deep technical specialization at the life sciences and compute interface. The increase in both fund size and pacing is notable given the general contraction in venture fundraising that has characterized much of the post-2022 period.
The broader context here is that what distinguishes Dimension's position is the dual-path exposure to both the scientific application layer, through companies like Chai Discovery and New Limit, and the compute infrastructure layer, through investments like Modal Labs and the resulting Anthropic equity. That structure means the firm benefits whether value accrues to the AI platforms themselves or to the scientific tools built on top of them. We have seen this pattern before, when firms that backed both infrastructure providers and the SaaS companies consuming that infrastructure during the cloud buildout of the 2010s captured compounding returns from adjacent layers of the same stack. Dimension appears to be applying a version of that strategy to the intersection of machine learning infrastructure and computational science.
The third fund provides the capital to extend that approach. What remains is execution: whether the firm can continue identifying companies that sit at the productive boundary of scientific research and compute, and whether the valuations implied by the Chai Discovery and New Limit rounds hold up as those companies mature. For now, the capital is in place.


