DeepSeek's $12 Billion Round Shows the Chip-Light Bet Paid Off

DeepSeek is reportedly closing a $12 billion round backed by Tencent and CATL before a 2027 IPO, well above the $7.4 billion it first targeted.

DeepSeek is reportedly close to raising $12 billion in a new funding round, according to Bloomberg's report as summarized by PYMNTS. The Chinese startup originally aimed for around $7.4 billion (50 billion yuan). Contemporary Amperex Technology Co. and Tencent are among the largest backers, and signed term sheets suggest the total could approach $15 billion. The round is expected to close soon, ahead of a listing planned for next year.

The reason given for the overshoot is demand: interest exceeded expectations after the release of the company's latest model. That is worth pausing on. A round that grows by roughly 60 percent over its target, before the company has even listed, tells us that investors are pricing the next model release as much as the current business. We'd treat that as a signal about sentiment, not a verified measure of what DeepSeek will earn.

Initial target versus reported round size, with term sheets pointing toward $15 billion.

Remember why DeepSeek mattered in the first place. Early last year it shipped a model with performance comparable to its U.S. rivals while using significantly fewer Nvidia chips, which rattled both Silicon Valley and Wall Street. This round is, in effect, the market paying for that efficiency story. It also helps that a report from The Information, cited by PYMNTS, says the company more than doubled its annualized revenue run rate in recent months, to $1 billion. That figure comes from unnamed sources, so we hold it loosely.

There is a competitive angle too. Moonshot is reportedly aiming for its own IPO in early 2027 after closing funding at a $50 billion valuation. Two Chinese AI labs heading to public markets within months of each other would give investors a rare side-by-side comparison, and it would put pressure on both to show revenue rather than benchmark wins. Bloomberg calls DeepSeek's listing one of the most anticipated debuts on the Chinese market in years.

Notice what the material does not tell us: the valuation, the use of proceeds, or how much of that $1 billion run rate is recurring. Those are the numbers that will decide whether this is a durable business or a well-timed moment. Our take is cautious. The story is real, the backers are serious, and the efficiency argument has aged better than most. But a round that exceeds its target is a statement about appetite, and appetite can change faster than training runs.

Opinion · My 12-month bet

I expect the IPO to happen on schedule and to price richly, because the demand described in the reporting is already there. The harder test comes after listing, when a $1 billion revenue run rate has to justify a valuation set by hype.

Within twelve months I expect Moonshot to follow, and I expect investors on both sides to ask the same question: is the efficiency edge durable, or was it a one-time surprise? I lean toward the second answer.

Related articles from LaRebelión

Mistral says its new model beats Chinese rivals on cyber, but shows no numbersSouth Korea bets $3.5 billion on its own frontier AI modelGoogle folds cloud migration agents into a single Cloud Modernize push

Original source: PYMNTS.com

Produced with AI support and reviewed by the newsroom

Falcon · Signals analyst · Riyadh

“Oversubscription tells us what investors believe, not what the model will earn once the IPO hype fades.”