A company raising $7.4 billion then immediately pausing its next round isn't hesitation—it's a stress test of what happens when AGI ambition collides with investor impatience.
The Summary
- DeepSeek paused second funding round discussions weeks after closing a $7.4 billion raise at $52 billion valuation, then reportedly targeted over 10 billion yuan ($1.4B) at a $71 billion valuation
- Conflicting valuations across three reported rounds in under a month suggest either chaotic cap table management or founders testing how much rope the market will give them
- Founder says AGI development, not profit, remains core mission while pledging to keep top models open-source, a stance that might explain why traditional investors are getting cold feet
The Signal
DeepSeek's funding trajectory looks less like a roadmap and more like a seismograph during an earthquake. The company closed a $7.4 billion round at a $52 billion valuation, then within weeks started pitching a follow-on for at least 10 billion yuan at a $71 billion valuation. Now those discussions are paused. That's a 37% valuation jump in the time most startups are still drafting their board deck.
The whiplash makes sense when you know what DeepSeek is actually selling. Or rather, not selling. The company's founder has publicly stated that AGI development, not profit, is the core mission, and they plan to keep their most advanced models open-source. That's a beautiful vision if you're building a research lab. It's a nightmare if you're an LP trying to model returns.
"AGI, not profit, remains the company's core mission."
Open-source AI at this scale creates a fascinating paradox. You're burning billions to train models, then giving the weights away for free. The playbook assumes you'll monetize through services, enterprise support, or infrastructure, but that's never been proven at DeepSeek's velocity. OpenAI has the luxury of a closed model and a $10 billion Microsoft backstop. Anthropic has similar constraints and similar backers. DeepSeek is trying to run the open-source play at closed-source burn rates.
The pause might be DeepSeek pumping the brakes. Or it might be investors realizing they're funding a public good with venture returns expectations. Either way, the gap between the initial $45 billion valuation reported just days earlier and the later $71 billion ask suggests someone is recalibrating what this company is actually worth if profit isn't the North Star.
Key tensions:
- $7.4B raised, another $1.4B+ targeted, then paused, all in under 30 days
- Valuation swings from $45B to $52B to $71B depending on which round you're reading about
- Open-source model strategy with closed-source capital requirements
The Implication
If DeepSeek's funding chaos tells us anything, it's that the market still doesn't know how to price AGI development that refuses to play the margin game. Investors want OpenAI-style returns. Founders want to build the future and open-source it. Those two things don't fit on the same cap table without someone getting hurt.
Watch what DeepSeek does next. If they restart the round at a lower number, it means reality won. If they hold firm or go higher, it means they found believers willing to fund a moonshot with no exit. Either outcome sets a precedent for every other AI lab trying to figure out whether "build AGI and give it away" is a fundable thesis or a fairy tale.