Accel in Talks to Lead $1B Round for Thinking Machines at $40B Valuation
Thinking Machines, founded by former OpenAI CTO Mira Murati, is in talks to raise $1 billion at a valuation of at least $40 billion The company's annual revenue run rate reportedly exceeds $100 million, implying an unusually high revenue multiple Accel is said to be leading the round, with the valuation falling below the $50 billion sought late last year The startup launched Inkling, an open-weight model generating revenue through usage-based compute fees via its Tinker platform Several high-pro
Analysis
TL;DR
- Thinking Machines, founded by former OpenAI CTO Mira Murati, is in talks to raise $1 billion at a valuation of at least $40 billion
- The company's annual revenue run rate reportedly exceeds $100 million, implying an unusually high revenue multiple
- Accel is said to be leading the round, with the valuation falling below the $50 billion sought late last year
- The startup launched Inkling, an open-weight model generating revenue through usage-based compute fees via its Tinker platform
- Several high-profile departures have occurred since the previous $2 billion seed round, including co-founders Lilian Weng and Luke Metz returning to OpenAI
Why It Matters
This round highlights the intense investor appetite for talent-driven AI startups, even those with elevated valuations and notable leadership turnover. The revenue multiple alone signals how the market is pricing future potential over current financials in the competitive AI landscape. For practitioners, it underscores the importance of platform strategies—Thinking Machines is monetizing through compute fees rather than traditional model licensing.
Technical Details
- Inkling: An open-weight model launched in July that enables customization on proprietary data through the Tinker platform, using a usage-based compute fee monetization model
- Tinker Platform: A customization infrastructure allowing users to fine-tune Inkling on their own data, representing a shift toward infrastructure-as-a-service in the AI model space
- Previous Fundraise: A $2 billion seed round led by Andreessen Horowitz with participation from Nvidia, GV, Lightspeed, and Conviction Partners, valuing the company at $12 billion
- Revenue Model: Usage-based compute fees rather than direct model sales, aligning with a platform-centric approach to AI monetization
Industry Insight
- The $40 billion valuation on sub-$100M revenue reflects the continued "bet on the jockey" mentality in AI investing, where founder pedigree and team composition outweigh traditional financial metrics
- High-profile departures (Weng, Metz returning to OpenAI) suggest that even well-funded AI labs face retention challenges, and investors should weigh team stability carefully when evaluating valuations
- The shift toward open-weight models with platform-based monetization (compute fees for customization) represents an emerging strategy to compete with closed-model giants while building recurring revenue through infrastructure services
Disclaimer: The above content is generated by AI and is for reference only.