DeepSeek is about to get a lot more money. The Chinese AI lab is close to raising at least $12 billion in fresh funding, and strong investor demand could push the round as high as $14.9 billion, according to Bloomberg reporting this week. Tencent and battery giant CATL are among the biggest backers lining up for a piece of it.
That would make it one of the largest private funding rounds in AI history, and it is not happening in isolation. Across the Pacific, OpenAI is seeking at least $30 billion at a $1.4 trillion valuation, Moonshot AI just closed a private round at a $50 billion valuation ahead of a Hong Kong IPO, and Anthropic is reportedly eyeing a November IPO at a valuation around $2 trillion. The AI funding frenzy is not slowing down. If anything, October is the month it went into overdrive.
The October funding scoreboard
Here is where the big rounds stand, based on this week’s reporting:
| Company | Raise / valuation | Key backers | Next step |
|---|---|---|---|
| DeepSeek | $12B to $14.9B raise | Tencent, CATL, state funds | IPO planned early 2027 |
| OpenAI | $30B+ at $1.4T valuation | BlackRock, UAE funds (up to $10B) | Raise in progress |
| Moonshot AI | $50B valuation (final private round) | Not disclosed | Hong Kong IPO Q1 2027, targeting up to $5B |
| Anthropic | ~$2T IPO valuation target | Public markets | IPO eyed for November |
For Indian readers, the DeepSeek number alone is staggering: $12 billion is roughly Rs 1.06 lakh crore at current exchange rates, and $14.9 billion works out to about Rs 1.31 lakh crore. That is more than the market value of most listed Indian companies, raised by a single AI lab in a single round.
Why investors are piling in now
The immediate trigger for the DeepSeek round is product momentum. The company recently released its V4 Flash model, which strengthened its standing against OpenAI and Anthropic on both cost and performance, and investor appetite grew enough to push the round past its original target of around $7.5 billion. When a lab can match frontier performance at a fraction of the price, the market pays attention.
The bigger story is compute. Training and serving frontier models is an infrastructure game now, and the labs are racing to lock in chips, data centers, and energy years in advance. DeepSeek’s raise, OpenAI’s $30 billion ask, and the state-backed funds circling the Chinese rounds are all really about the same thing: securing the physical capacity to train the next generation of models. Money is the moat.
Moonshot’s numbers tell the same story from the revenue side. The company behind the Kimi chatbot expects annual recurring revenue to hit $2 billion by December, double its current level. When a chatbot business can plausibly double revenue in months, a $50 billion valuation starts to look less like speculation and more like arithmetic, at least to the investors writing the checks.
What this means for model prices
Here is the part that matters if you actually use these models. DeepSeek built its reputation on undercutting everyone on price, and a $12 billion war chest lets it keep doing that. Every dollar of subsidized compute that flows into an open-weights lab is downward pressure on the per-token prices that OpenAI, Google, and Anthropic can charge.
We have been tracking those prices all month. The October 2026 AI price list shows frontier models spanning from a few cents to tens of dollars per million tokens, and the spread keeps widening as cheaper labs force the leaders to respond. OpenAI’s answer was GPT-6.1 Sol at $2 per million tokens; Google’s was Gemini 4 Argon at the same price point. DeepSeek raising $15 billion is a bet that the price war continues, and that it wins it.
The bubble question, honestly asked
No honest article about these numbers can skip the obvious question. Gartner forecasts global AI spending will reach $2.6 trillion this year. The S&P 500 just hit another all-time high near 7,830 points, with Nvidia alone worth close to $6 trillion and accounting for more than 8% of the index. Anthropic’s $2 trillion IPO target would value it at a multiple of its revenue that would have been unthinkable eighteen months ago.
The bull case is that the revenue is real this time. Moonshot doubling to $2 billion in ARR, enterprises actually deploying agents, and token prices falling far enough for mass adoption are all things you can point to. The bear case is that valuations are pricing in a decade of that growth arriving in two years.
What is not in doubt is the direction of travel. Whether these rounds look brilliant or reckless in hindsight, the next twelve months of AI progress are now fully funded. For the rest of us, that means cheaper models, faster releases, and a US-China race that is being fought with checkbooks as much as with chips.
