Labelbox, which develops data annotation and labeling software, raises $110M Series D led by SoftBank's Vision Fund II, says it is now “basically a unicorn”
With funding for data startups in a frenzy, data annotation company Labelbox found itself fielding offers … Tweets: @glenrdixon , @kenrickcai , and @firstround Tweets: Glen Dixon / @glenrdixon : True to its name, Labelbox is building for the data labeling part of the machine learning process, which enables smaller batches of “training data” to be annotated so that AI models can learn how to make predictions and develop insights from larger amounts of raw data. https://twitter.com/... Kenrick Cai / @kenrickcai : SoftBank's Vision Fund II is placing its bet on Labelbox, a data labeling startup in a space with already a couple $1 billion companies. Labelbox has raised $110 million, and its eyes are on cornering business in the enterprise and government. My story: https://www.forbes.com/... @firstround : Congrats to @manuaero, @Riegerb & the entire @labelbox team for announcing their Series D fundraise! Labelbox saves enterprises time by creating & managing AI training data, people & processes in 1 place (& good news—they're hiring!). Read on in @forbes: https://www.forbes.com/...
Context & Ripple Effects
Labelbox's raise caps a rapid climb through the AI tooling stack: a $10M Series A led by Gradient Ventures in 2019, then a $25M Series B with a16z and a $40M Series C with B Capital, before SoftBank's Vision Fund II writes the largest check yet. The Forbes reporting notes the company was fielding offers amid a frenzy of funding for data startups — the raise was competitive, not opportunistic.
The bet puts SoftBank's flagship growth fund behind the annotation layer rather than the models themselves, and it lands in a category where peers like CloudFactory ($65M in 2019) had already established that labeling-for-ML is a financeable business.
First-order effects
- Labelbox's total raised roughly triples overnight, from $79M to about $189M, giving the company war chest and 'basically a unicorn' status claim while rivals like CloudFactory remain sub-$100M funded.
- Vision Fund II adds a data-annotation platform to its portfolio at the moment its capital was actively seeking AI infrastructure exposure.
Second-order effects
- Competitors in the same budget line — CloudFactory on services-heavy labeling, and later Cleanlab with its $100M-valued data-quality Series A — now face a rival that can outspend them on product and land large enterprise contracts on price.
- Investors reading the frenzy take the hint that training-data tooling clears unicorn barriers, pulling more term sheets into the category.
Third-order effects
- If model quality keeps tracking labeled-data quality, the data layer consolidates into capitalized platforms the way cloud tooling did — with annotation shifting from outsourced labor lines to strategic software spend.
- The pattern points toward AI supply chains being financed end-to-end: compute, models, and now the training data beneath them each attract dedicated growth capital.
The trend: Venture capital is increasingly treating training-data tooling as core AI infrastructure, moving the money down the stack from models to the labeled data they depend on.