DeepSeek's $7.4 billion first-ever funding round, backed by Tencent Holdings and CATL, would place the Chinese AI lab in the same capital tier as OpenAI.
DeepSeek, the Chinese AI lab that rattled Western competitors in early 2025 by releasing large language models that matched frontier performance at a fraction of the reported training cost, is preparing to raise approximately 50 billion yuan -- around $7.4 billion -- in its first-ever external funding round. Reuters first reported the deal, which would rank among the largest single fundraises in the global artificial intelligence sector.
Tencent Holdings and electric vehicle battery maker CATL are expected to participate, according to people familiar with the matter. The round has not been officially confirmed by DeepSeek.
The investor lineup
CATL's presence is the more surprising entry. The company dominates global EV battery supply chains but is not an obvious technology investor. As artificial intelligence infrastructure scales up -- demanding more power, specialized hardware, and tighter integration with physical manufacturing -- industrial firms are increasingly seeking early positions in AI development. Backing a frontier lab is one way to ensure access, partnership, or influence as that convergence accelerates.
For Tencent, the logic is more conventional. The company operates at scale across social platforms, gaming, cloud, and enterprise software, and has already seeded a broad range of Chinese AI startups. Adding DeepSeek to its portfolio would give it a stake in the lab widely regarded as China's most technically rigorous.
DeepSeek's rise
DeepSeek attracted sustained international attention in early 2025, when benchmark results placed its models competitively against systems from OpenAI and Google, despite operating under U.S. semiconductor export restrictions that limit access to advanced Nvidia GPUs. The inference efficiency of its models challenged the prevailing assumption that frontier AI requires ever-larger capital budgets -- a debate that remains unresolved, but one DeepSeek gave serious empirical weight.
American Bazaar reported that the prospective fundraise reflects growing confidence among Chinese tech and industrial firms in the domestic AI ecosystem. Beijing has made artificial intelligence a central policy priority amid ongoing chip export controls, pushing investment toward homegrown model development.
The $7.4 billion figure lands in a week when AI funding activity is particularly dense. Generative music startup Suno announced a $400 million Series D at a $5.4 billion valuation, according to SiliconAngle -- its second major raise in under six months. At roughly 5 percent of DeepSeek's reported target, it illustrates how much capital is concentrating at the frontier model tier.
The competitive stakes
OpenAI remains the benchmark. By early 2026, the company had reached an $830 billion valuation, with Forbes citing $20 billion in annual recurring revenue and 800 million monthly active ChatGPT users. Anthropic and xAI have also raised billions over the past two years. A DeepSeek close at $7.4 billion would install a Chinese lab at that same funding tier, without the years of prior external investment that built its Western rivals.
DeepSeek has operated as a research unit inside quantitative hedge fund High-Flyer, self-funded until now. Google, meanwhile, continues rapid model iteration: Android Authority reported this week on another Gemini 3.5 Flash update targeting improved performance on demanding software engineering tasks. The pace of deployment on both sides makes clear that no lab can afford to stand still.
What changes when a lean, research-focused lab accepts billions in outside capital is not a trivial question. Investor expectations tend to pull organizations toward commercialization, enterprise contracts, and recurring revenue -- away from the open model releases and benchmark-focused research that built DeepSeek's reputation. Whether that tension surfaces here depends on who gets board seats and what commitments accompany the capital.
The round is unconfirmed and terms remain undisclosed. But if Tencent and CATL are in, this is less a funding event than a declaration -- that China's largest technology and industrial companies have decided which AI lab they are backing in the next phase of the race.
Frequently asked questions
What is DeepSeek?
DeepSeek is a Chinese AI research lab affiliated with quantitative hedge fund High-Flyer. It gained global recognition in early 2025 after releasing large language models that performed competitively on standard benchmarks against Western systems, reportedly at significantly lower training cost.
Who is investing in the round?
Tencent Holdings and battery manufacturer CATL are expected to participate, according to Reuters. No terms have been publicly confirmed and DeepSeek has not commented officially.
How does this compare to OpenAI's funding?
OpenAI held an $830 billion valuation in early 2026 with $20 billion in annual recurring revenue. DeepSeek has no public revenue figures and has primarily released open-weight models. A $7.4 billion raise would match Western rivals in fundraising scale, though not yet in commercial deployment.
Does outside investment change how DeepSeek operates?
Possibly. The lab has run as a lean, research-focused unit with no prior external investors. Large raises typically bring pressure to build enterprise products and prioritize revenue -- factors that have reshaped several other AI labs after taking on major outside funding.