How much has Scale AI raised?

Scale AI raised approximately $1.6 billion in venture equity across roughly six priced rounds from 2016 through its May 2024 Series F, which valued the company at about $13.8 billion. Those rounds were straightforward equity financings — there is no large publicly reported debt facility, because Scale is a services-and-software business rather than a balance-sheet lender.

The much larger number is Meta's June 2025 investment of $14.3 billion for a ~49% non-voting stake. That capital went largely to liquidity for existing shareholders rather than onto Scale's balance sheet the way a normal primary round would, which is why it is best understood as a strategic investment plus partial exit, not a $14B 'raise' to fund operations. Scale had already tested the secondary market with a ~$25B tender offer earlier in 2025 before the Meta deal superseded it.

Scale's cap table reads like a roster of top venture and growth funds. Accel led the Series A and the 2024 Series F and was Scale's single largest outside holder; Founders Fund (Peter Thiel) led the 2019 unicorn round; Index Ventures led the Series B; and Tiger Global, Dragoneer, and Greenoaks led the 2020–21 growth rounds. NVIDIA and Amazon joined as strategic investors in 2024, reflecting Scale's centrality to the AI supply chain.

Meta is now the dominant economic owner with its ~49% non-voting position. The non-voting structure matters: Meta gets the financial upside and a strategic relationship, but Scale's board and management retain control — a deliberate design so the company can keep selling to enterprises and governments without being treated as a Meta subsidiary.

Scale's valuation roughly doubled at each major step — $1B (2019), ~$3.5B (2020), ~$7B (2021), ~$13.8B (2024), and ~$29B (2025) — tracking the explosion in spending on training data and human feedback as frontier labs raced to build better models. Unlike many 2021-era startups, Scale never took a down round; demand for high-quality data kept rising through the LLM boom, and the company kept doubling revenue (to ~$2B in 2025).

The 2025 step to ~$29B was driven less by organic financials than by strategic scarcity: Meta wanted Scale's data engine and, above all, its founder and top researchers. The premium reflects an acqui-hire-plus-investment dynamic. The risk on the other side is that the same deal drove rival labs (Google, OpenAI, Microsoft, xAI) to pull work, pressuring the very revenue base that justified the price — though 2025's doubled revenue suggests the enterprise and government rebuild largely offset the lab churn.

Scale has not disclosed audited profitability. With an estimated 50%+ gross margin and revenue roughly doubling to ~$2 billion in 2025, the unit economics are workable but lower than pure software because of the heavy human-labor component. The July 2025 layoffs (14% of staff plus ~500 contractors) were framed by interim CEO Jason Droege as removing scaled-up bureaucracy in commodity labeling and refocusing on enterprise, government, and evaluation.

An IPO is not imminent. Meta's ~49% stake and the loss of several frontier-lab customers make a near-term public offering complicated; the more likely path is a multi-year rebuild around enterprise, government, and model-evaluation revenue before any liquidity event. Investors already received substantial cash from the Meta transaction and the earlier tender, which reduces pressure to exit via IPO.

For a revenue team, the Meta deal is a mixed but informative signal. On one hand, Scale is extremely well-capitalized and is standing up new motions in enterprise 'Apply AI', government, and evaluation — each a potential entry point with fresh budget owners who need tooling, infrastructure, and services, and 2025's record bookings mean money is moving.

On the other hand, the 2025 customer exits and layoffs mean budgets are being scrutinized and reorganized, so timing and the specific buying center matter more than usual. Sell to the team being funded for growth (government, evaluation, GenAI platform) rather than the contracting data-labeling org, and expect a more security- and procurement-conscious buyer given Scale's government work, its Meta relationship, and an active U.S. Department of Labor review of its labor practices.