Subscribe
News
4 min read

Fal Raises $1.65B; Bloomberg Reports $16B Valuation

Fal announced $1.65B across a Series E and F. Bloomberg reported the $1.3B Series F values it at $16B, double April's $8B.

Olivetti-style cover on cream paper with red, mustard and teal blocks, thick charcoal rules, a flat GPU rack icon and the title FAL $1.65B RAISE in bold sans type.
Olivetti-style cover on cream paper with red, mustard and teal blocks, thick charcoal rules, a flat GPU rack icon and the title FAL $1.65B RAISE in bold sans type.

Fal, a San Francisco company that hosts image, video and audio AI models, announced on Tuesday, November 3, that it has raised a combined $1.65 billion across a Series E and a Series F. The law firm that advised fal, Ashurst Perkins Coie, says Bessemer Venture Partners led the $1.3 billion Series F with continued investment from NVIDIA, and Sequoia Capital led the Series E. The valuation comes from Bloomberg, which is behind a paywall: as relayed by SiliconANGLE, Bloomberg reported that the Series F values fal at $16 billion.

Fal's own announcement post names investors and uses but states no dollar amounts, so each figure below names the outlet that states it.

What was announced

The Series E closed earlier this year and was announced together with the Series F. Per Ashurst Perkins Coie, the Series E included significant investment from Kleiner Perkins and a new investment by NVentures. SiliconANGLE, citing Bloomberg, reports a $350 million Series E at an $8 billion valuation. $1.3 billion plus $350 million equals the $1.65 billion combined figure.

Per SiliconANGLE, Sound Ventures, HOF Capital and T. Rowe Price Group also joined the Series F, fal described NVIDIA as a strategic investor, and Bessemer partner Talia Goldberg is joining the board of Features & Labs Inc., as fal is formally known.

What the money buys

Fal's post says the work takes two things: "the inference capacity to run these models at scale and the team to build what comes next." It adds that the rounds "fund work across the fal ecosystem in 2027" and give the team "room to pursue new research and new forms of creation."

SiliconANGLE, citing Bloomberg, adds that the money goes to computing power and to hiring beyond fal's current 185 employees. Co-founder and CEO Burkay Gur told Bloomberg: "We hit a point in summer where we were not able to serve our customers and had to rate-limit them." Per the same report, pretraining a model of fal's own is not on the near-term agenda, and Gur did not rule it out for later.

Valuation path

Fal's valuation has risen in each round reported by Bloomberg:

  • December 2025: a $140 million Series D at $4.5 billion, per TechCrunch, citing Bloomberg.
  • April 2026: a $350 million Series E at $8 billion, per SiliconANGLE, citing Bloomberg.
  • Series F: $1.3 billion at $16 billion, per SiliconANGLE, citing Bloomberg.

The division: $16 billion is 2.0 times the April figure, which matches TechCrunch's word "double," and about 3.6 times the December figure.

Bars to scale from zero showing fal valuation of $4.5B, $8B and $16B and round sizes of $140M, $350M and $1.3B

Revenue and peers

SiliconANGLE reports that fal's annualized revenue passed $800 million earlier this year, a figure Bloomberg attributed to a person familiar with the matter. Fal's announcement post does not state it.

Two other inference companies have raised large rounds this year:

  • Fireworks announced a $1.505 billion Series D at a $17.5 billion valuation on July 15, and said its annualized revenue run rate passed $1 billion, in its own post.
  • Modal announced a $750 million Series D at a $15.75 billion post-money valuation on October 14, and said it crossed $1 billion in annualized run rate, in its own post.

Simple division gives about 20 times for fal ($16 billion over $800 million), 17.5 times for Fireworks and 15.75 times for Modal. Treat these as ceilings, not a ranking. Each revenue figure is a floor ("passed," "crossed"), the dates differ, and the reports do not label fal's $16 billion pre-money or post-money.

TechCrunch adds a caution for the sector. Several people familiar with inference companies' financials told it that gross margins stay thin because compute is expensive to buy or lease. That comment covers the sector, not fal alone.

Fal's models and partners

Bessemer, in its post on leading the round, says Adobe, Canva, Shopify, World Labs and Amazon have moved production volume to fal. It also says fal Research released H3 Max, a post-trained version of MiniMax H3, in August, and partnered with Black Forest Labs on Flux 3 Max. Our FLUX 3 Max explainer covers that model's speed claim, rankings and pricing.

SiliconANGLE, citing Bloomberg, adds that fal's in-house researchers post-train some partner models to run faster, and that some of those deals pay fal licensing fees or a share of revenue.

Why it matters

This round is funding for inference capacity. It sits next to a report that a Fluidstack round would value the neocloud at $40B and Nvidia's talks to back Nscale debt. The business models differ, but all three concern funding for compute.

For fal, the stated use of funds is compute and people. The demand evidence in the reports is Gur's rate-limit comment and the $800 million figure, and Bloomberg is the source for both. The figure comes from an unnamed person.

What to watch

NVIDIA, which fal names as a Series F investor, has set its next dated event. Per its investor relations release, NVIDIA will host a conference call on Tuesday, November 17, at 2 p.m. PT to discuss results for its third quarter of fiscal year 2027, which ended October 25, 2026.

Sources

NewsFundingInfrastructureInference