Key Notes
- DeepSeek is nearing at least $12 billion in new funding, with Tencent and CATL reportedly committing among the largest amounts in the round.
- The company targeted a valuation of roughly $75 billion, while the final financing terms and a planned early-2027 IPO remain unsettled.
- The proposed investment comes as AI developers seek capital for expansion while working to lower the cost of running their models.
DeepSeek is close to securing at least 80 billion yuan, or about $12 billion, in a new funding round, Bloomberg reported on October 6, citing people familiar with the matter. The financing would give the Chinese artificial intelligence developer additional capital ahead of a planned stock-market debut.
Tencent Holdings Ltd. (HKEX: 0700) and battery manufacturer Contemporary Amperex Technology Co. Ltd., known as CATL (HKEX: 3750), have committed among the largest amounts, according to the report. DeepSeek had targeted a valuation of approximately 500 billion yuan, or roughly $75 billion.
That valuation is a fundraising target, rather than the market capitalization of a listed company. The round has not been announced as completed, and the final terms could change.
Investor Demand Pushes the Round Beyond Its Target
DeepSeek initially sought about 50 billion yuan. Signed term sheets could take the final amount toward 100 billion yuan, Bloomberg reported. Its sources linked stronger investor interest to the company’s recent model releases.
The minimum reported amount is already 60% above that original fundraising goal. A larger round would increase financial flexibility, but the amount raised and the valuation assigned to the business are separate measures. The reported terms do not establish a final ownership percentage for either Tencent or CATL.
Tencent’s AI Spending Provides a Financial Backdrop
Tencent’s own quarterly results show how heavily it is investing in AI. The company reported capital expenditure of 52.8 billion yuan in the second quarter of 2026, up 176% from a year earlier, as it expanded computing capacity.
It also reported negative free cash flow of 13.8 billion yuan. Tencent said large AI-related compute prepayments contributed to the cash outflow; excluding those prepayments, free cash flow would have been positive at 37.6 billion yuan.
The distinction illustrates why the timing of infrastructure payments matters alongside revenue growth. AI companies can commit substantial cash before the computing capacity supports the sales expected from it. Tencent identified its own models, WorkBuddy, CodeBuddy, Weixin initiatives and cloud demand among the uses of that capacity.
Those figures describe Tencent’s operations, not its proposed investment in DeepSeek. They provide context for the reported financing without establishing how much Tencent will contribute to the round.
CATL Brings the Energy Side of the AI Expansion
CATL’s involvement connects the funding story with the physical infrastructure supporting AI. The company’s businesses include electric-vehicle batteries and stationary energy-storage systems, placing it in a different part of the technology supply chain from an internet-platform operator.
In its June storage announcement, CATL described growing AI-related electricity demand as one reason energy storage is becoming critical infrastructure. It introduced a sodium-ion storage system and outlined plans to expand production and deliveries.
That strategy helps explain the commercial relevance of AI infrastructure to battery suppliers. It does not establish a battery-supply agreement with DeepSeek, preferential contracts or any other operating arrangement attached to the reported investment.
Model Efficiency Remains Central to DeepSeek’s Business
DeepSeek’s September 10 model announcement introduced V4.1 Flash with native visual understanding and an emphasis on faster inference and lower operating costs. The company said its architecture reduces the computing work required to process and generate responses.
DeepSeek also said the model cuts requirements for high-bandwidth memory and solid-state storage to one-quarter and one-eighth, respectively, compared with its predecessor’s context-storage approach. These are company-reported technical claims, rather than independently verified reductions in its overall expenses.
The commercial implication is that serving more customer requests with fewer resources could improve the economics of an AI service. DeepSeek paired the release with lower API pricing and retained a peak-and-off-peak schedule, with off-peak prices at half the peak rate.
Efficiency therefore creates a balancing act: lower costs can support cheaper services and wider adoption, while lower prices mean the company must also build sufficient usage to sustain revenue.
The Round Joins a Wider Race for AI Capital
The financing would place DeepSeek among the AI developers seeking large pools of private capital before a public listing. MarketSpeaker’s OpenAI coverage examines another proposed round, targeting at least $30 billion at a reported $1.4 trillion pre-money valuation.
Those transactions involve different companies and valuation bases, so their headline figures should not be treated as directly comparable ownership prices. Both nevertheless show why access to financing remains a major issue even for developers focused on reducing the cost of AI.
Bloomberg reported that DeepSeek is preparing for an IPO in early 2027. The next developments to watch are completion of the private round, disclosure of its final valuation and the formal steps toward a listing. Until then, the funding amount and IPO timetable remain reported plans.
