DeepSeek nears massive funding round with potential valuation near $15 billion

What risks come with such rapid valuation growth?

Chinese AI startup DeepSeek is reportedly in advanced talks to raise over $12 billion in a new funding round that could push its valuation close to $14.9 billion, according to sources familiar with the matter. The round, which may close soon, would mark one of the largest private investments in the artificial intelligence sector to date. Major Chinese technology and industrial firms, including Tencent and Contemporary Amperex Technology Co. Limited (CATL), are expected to be the lead investors. The funding surge reflects growing confidence in DeepSeek’s rapid progress in developing large language models capable of handling extremely long-context DeepSeek’s technological edge lies in its Subquadratic architecture, designed to process inputs of up to 12 million tokens in a single pass without relying on retrieval-augmented generation workarounds.

This capability allows the model to analyze entire codebases, legal documents, or research archives cohesively, setting it apart from conventional LLMs that struggle with long-range dependencies. The company’s SubQ 1.1 Small variant has demonstrated near-perfect retrieval accuracy across extended contexts, a breakthrough that could transform how enterprises manage and query vast internal data sources. Industry analysts note that such performance could reduce infrastructure costs and improve efficiency in sectors like software development, finance, and legal tech.

Frequently Asked Questions

Despite the excitement, some experts caution that DeepSeek’s soaring valuation may outpace its current revenue streams and real-world deployment scale. The AI sector remains volatile, with regulatory scrutiny increasing in both China and Western markets over data security and model transparency. Additionally, reliance on major corporate investors like Tencent and CATL could raise concerns about strategic influence or geopolitical tensions affecting international partnerships. DeepSeek will need to demonstrate sustainable monetization and ethical AI practices to justify its lofty valuation in the long term.

What makes DeepSeek’s SubQ model different from other LLMs? Its Subquadratic architecture allows it to process up to 12 million tokens in a single pass without RAG, enabling coherent Why are Tencent and CATL investing in DeepSeek? Tencent brings cloud and ecosystem scale, while CATL offers industrial AI applications in energy and manufacturing, suggesting a dual focus on consumer tech and enterprise infrastructure.

Is DeepSeek planning to release its models publicly? The company has not announced open-source plans, but it may offer API access or enterprise licenses as part of its commercialization strategy following the funding round.