Mira Murati's startup lands $2B at a $12B valuation

If you want to know what Silicon Valley is willing to throw blank checks at right now, look no further than ex-OpenAI leadership. You don’t even need a public product or a demo tape yet—just assemble the right brains, and billions will follow.

Former OpenAI CTO Mira Murati just confirmed a massive $2B funding round for her stealth startup, Thinking Machines Lab. The round was led by venture powerhouse a16z, with strategic heavyweights including NVIDIA, AMD, Cisco, Accel, ServiceNow, and quantitative trading firm Jane Street all pitching in.

The fresh cash injection instantly catapults the startup to a staggering $12B valuation. That is a massive milestone for a company founded just this past February, which has stayed almost entirely under the radar until now.

Multimodal collaboration and an open-source twist

Murati used the funding announcement to break her silence on what Thinking Machines Lab is actually building.

The lab's overarching mission is advancing what she calls collaborative general intelligence. Basically, they want to step beyond the standard single-prompt chat box and build multimodal systems that behave like real coworkers. That means AI that processes conversation, visual inputs, and the messy, iterative ways humans naturally collaborate.

The biggest headline for builders: Murati promised their first product will ship within the next couple of months. Even better, it will include a significant open-source component specifically tailored to help researchers and startups develop custom models.

She also noted that the lab will publish deep research to help the scientific community understand frontier systems, leaning heavily into transparency that many proprietary labs have abandoned.

Custom models and real-world cash

While Murati’s public pitch leans into research and collaboration, her commercial roadmap looks decidedly practical.

According to earlier reporting from The Information, Thinking Machines Lab is planning to develop custom AI models for enterprise clients aimed directly at driving bottom-line profits. It is a smart dual approach: win developer hearts and minds with open-source tools, while closing high-dollar enterprise contracts behind the scenes.

The lab has faced no shortage of acquisition rumors or investor hype over the past six months, but this round confirms Murati intends to stay independent and build an empire of her own.

Why it matters

We think this launch timeline puts serious pressure on the AI ecosystem in a couple of distinct ways.

For developers, the promise of a major open-source component from former OpenAI brains is huge. Open multimodal models have lagged behind closed-source giants like GPT-4o and Gemini. If Murati’s team delivers accessible weights or meaningful tooling that makes custom training practical, it could instantly become the default foundation for thousands of AI startups.

For the broader industry, the battle among former OpenAI executives is getting wild. Between Sam Altman’s OpenAI, Ilya Sutskever’s Safe Superintelligence, and now Murati’s $12B Thinking Machines Lab, the frontier landscape is fragmenting into well-funded rival factions. Heavy hitters like NVIDIA and AMD hedging their bets across all of them proves just how uncertain the ultimate winner remains.

That said, a $12B price tag before shipping line one of public software leaves zero room for error. When that first release drops in a couple of months, it cannot just be another conversational wrapper. It will need to demonstrate genuinely fresh collaboration dynamics to justify the staggering hype.