Harvey hits $15.5B valuation, months after reaching $11B

Harvey raised $550 million at a $15.5 billion valuation, the legal AI startup announced on Sept. 9, 2026. The new round, co-led by Diffusion and Lightspeed Venture Partners, follows earlier financings that valued the company at $11 billion in March and $8 billion in December.

By AI NewsroomPublished 42 minutes agoUpdated 42 minutes ago0 views

Harvey hits $15.5B valuation, months after reaching $11B

Why It Matters

The fundraising and product push signal strong investor appetite for legal-focused AI and back a model in which firms build on open-weight foundations rather than relying exclusively on frontier proprietary models. Harvey’s move to offer an in-house model and encourage customers to post-train open-weight weights highlights an alternative commercialization path for AI in regulated industries like law.

Key Facts

Funding amount: $550 million
Post-money valuation: $15.5 billion
Round co-leads: Diffusion and Lightspeed Venture Partners
Previous recent round: $200 million at $11 billion valuation (March 2026)
Earlier valuation: $8 billion (December 2025)

Harvey announced a $550 million financing at a $15.5 billion valuation on Sept. 9, 2026, marking another large infusion for the legal AI startup. The round was co-led by Diffusion and Lightspeed Venture Partners. Company disclosures show that Harvey has now raised more than $1.55 billion in total capital.

The new valuation continues a rapid climb: the firm reported a $11 billion valuation in March and an $8 billion valuation in December, representing a near doubling of value in roughly nine months. PitchBook estimates the company has completed at least eight priced rounds since 2023, five of them since 2025; because some of those were extensions, this latest financing could be classified as a Series F.

On the product side, Harvey recently introduced an in-house model called Harvey Tenet. The model was built from the open-weight Kimi K3 and then post-trained on legal data with assistance from inference provider Fireworks — the same provider that helped Cursor train its homegrown model. Alongside offering a proprietary model, Harvey is encouraging customers to adopt and post-train their own open-weight models.

Harvey’s approach illustrates a pathway in which a heavily regulated, domain-specific industry like law can make extensive use of AI while leaning on open-weight foundations instead of depending solely on proprietary frontier labs. The company did not name the financing round publicly; the announcement was posted Sept. 9, 2026.