Thinking Machines Funding: $1B Round at $40B Valuation

TL;DR: Thinking Machines seeks a $40 billion valuation in a $1 billion funding round led by Accel, implying a staggering 400x revenue multiple based on its $100 million annual run rate from the Tinker platform. This aggressive pricing highlights investor confidence in AI infrastructure despite high-profile founder departures and market scrutiny over sustainability.

Key facts

• Thinking Machines is in talks to raise $1 billion at a minimum $40 billion valuation, led by existing investor Accel.
• Founded in early 2025 by former OpenAI CTO Mira Murati, the company previously closed a $2 billion seed round in July 2025 at a $12 billion valuation.
• The proposed deal implies a ~400x revenue multiple based on an annual run rate exceeding $100 million from its Tinker platform.
• Thinking Machines positions itself as an AI infrastructure provider, generating revenue through usage-based compute fees for fine-tuning models via its Tinker platform.
• The company introduced Inkling in July 2025, an open-weight foundation model that serves as the base for enterprise customizations on the Tinker platform.
• Previous backers include Andreessen Horowitz (a16z), Nvidia, GV, Lightspeed Venture Partners, Conviction Partners, ServiceNow Ventures, NVentures, AMD Ventures, Cisco Investments, and Jane Street.
• Co-founders Lilian Weng and Luke Metz have departed to return to OpenAI, though other early researchers like John Schulman and Barret Zoph remain associated with the founding cohort.

The math behind a $40 billion AI valuation

Thinking Machines, the artificial intelligence laboratory founded by former OpenAI CTO Mira Murati in early 2025, is currently in discussions to raise $1 billion in its latest funding round at a minimum valuation of $40 billion [1][2][3]. This proposed figure represents a significant jump from the company’s previous seed valuation but introduces new questions about how investors are pricing early-stage AI infrastructure.

The capital injection is expected to be led by existing investor Accel, according to reports citing The Information and TechCrunch [1][3]. The startup had reportedly sought a $50 billion valuation late last year, suggesting that the current $40 billion ask reflects some adjustment in market expectations or negotiation dynamics [1][2][4].

Revenue multiples under scrutiny

The core of the debate surrounding this round lies in the financial metrics. Thinking Machines has achieved an annual revenue run rate exceeding $100 million [5]. At a $40 billion valuation, this implies a revenue multiple of approximately 400x [5]. Sources describe this multiple as extraordinarily high, even for the current AI boom, where valuations have been driven more by potential than proven profitability.

This pricing strategy hinges on Thinking Machines’ business model. Unlike many peers that compete directly on base large language models (LLMs), Thinking Machines positions itself as an infrastructure provider. Its revenue is generated through its Tinker platform, which allows users to fine-tune large models on proprietary datasets. The company charges usage-based compute fees for this service [1][2][3].

This model was anchored by the introduction of Inkling in July, an open-weight foundation model that serves as the base for these customizations [1][4]. By focusing on customization and infrastructure rather than competing for general-purpose model supremacy, Thinking Machines aims to capture value from enterprises seeking tailored AI solutions. The new $1 billion round is intended to scale compute infrastructure, expand its developer ecosystem, and accelerate enterprise product distribution [5].

A roster of heavyweight backers

The company’s financial backing has always been notable. Its previous fundraise was a massive $2 billion seed round closed in July 2025, which valued the startup at $12 billion [6]. That round was led by Andreessen Horowitz (a16z) and included participation from major tech and investment firms such as Nvidia, GV (Google Ventures), Lightspeed Venture Partners, Conviction Partners, ServiceNow Ventures, NVentures, AMD Ventures, Cisco Investments, and Jane Street [1][2][3][4][5].

Investor interest in the initial round was largely driven by the pedigree of Murati and the group of former OpenAI researchers who joined her. However, the company has experienced high-profile departures since its founding. Co-founders Lilian Weng and Luke Metz have left to return to OpenAI [1][2][3]. Other early employees associated with the founding cohort include John Schulman and Barret Zoph [6]. Despite these exits, investor confidence appears to remain strong enough to support a valuation jump of more than three times its previous worth.

Market context and speculation

The $40 billion valuation places Thinking Machines among the most valuable private AI startups globally. For comparison, established players like Anthropic have seen valuations fluctuate in similar ranges during their funding rounds, though they often have longer track records of revenue generation and broader product ecosystems [Editorial Note: Contextual comparison based on general market knowledge, not specific source data].

The current round highlights a divergence in how the AI sector is being valued. Some investors are betting heavily on infrastructure and developer tools, while others focus on application-layer companies or foundational model providers. Thinking Machines sits at the intersection of these categories, offering an open-weight model but monetizing through compute services.

Neither Accel nor Thinking Machines immediately responded to requests for comment regarding the new round [1][2][3]. As negotiations continue, the market will be watching closely to see if the $40 billion price tag holds or if further adjustments are made before the deal closes. The outcome of this round could set a benchmark for how future AI infrastructure startups are valued in an increasingly competitive landscape.

The focus on Tinker and Inkling suggests that Thinking Machines is betting on a future where customization and specialized compute power are more valuable than generic model access. Whether the 400x revenue multiple can be sustained as the market matures remains to be seen, but for now, investors seem willing to pay a premium for this specific vision of AI infrastructure.

Sources

1. Accel reportedly in talks to lead $1B round for Thinking Machines at $40B valuation | TechCrunch (techcrunch.com) — 2026-09-03
2. Accel in Talks to Lead $1B Round for Thinking Machines at $40B Valuation (theaiinsider.tech) — 2026-09-04
3. Accel reportedly in talks to lead $1B round for Thinking Machines at $40B valuation | TechCrunch (techcrunch.com) — 2026-09-03
4. Thinking Machines — Valuation, Funding & Investors | Value Add Intelligence (valueaddvc.com) — 2026-09-03
5. HyperAI (hyper.ai) — 2026-09-03
6. Mira Murati’s Thinking Machines Lab is worth $12B in seed round | TechCrunch (techcrunch.com) — 2025-07-15