Yes—but the headline needs one important qualification. Thinking Machines Lab, the artificial-intelligence startup led by former OpenAI CTO Mira Murati, was reported to have raised close to $2 billion in 2025 at an approximately $10 billion valuation before the investment. That implies a roughly $12 billion valuation after the new capital was added.
In other words, the commonly reported $10 billion and $12 billion figures can describe the same financing round. The difference is whether the source is referring to the company’s pre-money or post-money valuation.
As an Amazon Associate I earn from qualifying purchases.
The short version
Bloomberg reported on June 23, 2025, that Thinking Machines Lab had raised close to $2 billion at a $10 billion valuation before the investment. Andreessen Horowitz was reported to be leading the round, with Accel and Conviction Partners among the participants.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Later reporting attributed to a company spokesperson described the financing as $2 billion at a $12 billion valuation. The 2026 Stanford AI Index also records the round at $2 billion and a $12 billion valuation.
#1 Best Overall
The cleanest description is therefore:
Thinking Machines Lab raised approximately $2 billion in 2025 at an approximately $10 billion pre-money valuation, implying a roughly $12 billion post-money valuation.
Those numbers are approximate. The original report said “close to $2 billion,” and private financings can include terms, securities, options or other arrangements that make simple arithmetic less exact.
Why are both $10 billion and $12 billion reported?
The answer is standard venture-financing terminology.
Free tools Windows power users keep installed
One-click scans. No signup required.
| Term | Meaning in this transaction |
|---|---|
| Pre-money valuation | The company’s approximate value immediately before the new investment: $10 billion. |
| New capital | The amount investors put into the company: close to or approximately $2 billion. |
| Post-money valuation | The approximate value immediately after the investment: $12 billion. |
The basic calculation is:
$10 billion pre-money valuation
+ $2 billion new capital
= approximately $12 billion post-money valuationSo the $12 billion figure does not necessarily indicate a second funding round or a correction of the $10 billion figure. Bloomberg explicitly described the $10 billion valuation as being before the investment, which is effectively a pre-money valuation. A later source using $12 billion was referring to the approximate post-money value.
Did the round actually close?
The initial Bloomberg article described a financing in cautious terms, reporting what people familiar with the matter said about the amount, valuation and investors. It did not amount to a detailed public financing announcement from Thinking Machines Lab.
Later reporting attributed to a company spokesperson described the completed financing as $2 billion at a $12 billion valuation, and the Stanford AI Index retrospectively recorded the same figures. That supports describing the financing as having been completed, while still attributing the details rather than presenting them as terms published in a full company press release.
The available public record does not clearly establish a conventional Series A label or provide a complete term sheet. It is safer to call the event a private financing round than to assign it a specific series designation without a source.
Who invested?
Bloomberg reported that Andreessen Horowitz led or was leading the round, with Accel and Conviction Partners among the participants.
Later accounts associated with the completed financing also named major strategic and institutional investors including:
- NVIDIA
- AMD
- Cisco
- Jane Street
- ServiceNow
- Accel
- Andreessen Horowitz
This should not be treated as an exhaustive investor list. Private funding reports often identify only selected participants, and the company’s public news archive does not appear to contain a dedicated financing announcement listing every investor or each investor’s allocation.
Who is Mira Murati?
Mira Murati is the former chief technology officer of OpenAI. She co-founded and leads Thinking Machines Lab alongside other former frontier-AI researchers.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →At the time of the financing, the company’s valuation was tied primarily to the perceived strength of its founding team, its ability to recruit experienced AI researchers, its potential access to computing resources and investor expectations about the market for advanced, customizable AI. It was not based on publicly disclosed revenue or profits.
The financing also does not mean Murati personally received $2 billion. The money was raised by the company. A financing generally gives investors an ownership stake or related rights and can dilute existing shareholders, but the public information available here does not establish Murati’s ownership percentage, personal wealth or the precise allocation of shares.
What was Thinking Machines Lab building in 2025?
During the fundraising period, Thinking Machines Lab was highly secretive and had not publicly explained a finished product in detail. Contemporary reporting described its plans as unclear to the public.
Investor-facing descriptions reportedly emphasized customized AI for business customers and models tuned to organizational performance goals, according to The Information. That was reported information about what investors were told, not the same as a public product specification.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsThis distinction matters. In 2025, it was reasonable to describe Thinking Machines as a secretive or pre-product startup. It is no longer accurate to leave the company frozen in that description.
What has the company released since the financing?
Tinker: a model-customization platform
Thinking Machines announced Tinker on October 1, 2025. It is a platform and API for fine-tuning and post-training open-weight models.
Tinker exposes primitives such as forward_backward and sample, while the company’s Tinker Cookbook provides implementations of post-training methods. The service became generally available on December 12, 2025, removing its waitlist. Subsequent updates added OpenAI API-compatible sampling, vision input and support for additional models including Kimi K2 Thinking, according to the company’s general-availability announcement.
Tinker is aimed at researchers, developers and organizations that want to customize models rather than simply call a general-purpose chatbot. Its terms and service materials describe offerings that can include APIs, dashboards, model hosting, training, fine-tuning, evaluation, inference, storage and related tools.
The product uses usage-based pricing quoted per million tokens, with model prices varying by model and token category. The product page lists checkpoint storage at $0.10 per gigabyte-month. Prices and availability can change, and the company’s terms note that some services may be beta or subject to revision. Readers should check the Tinker product page and official model pricing documentation before making a purchasing decision.
Inkling: an open-weights model
On July 15, 2026, Thinking Machines announced Inkling, an open-weights model. The company describes it as a generalist system with reasoning, multimodal capabilities, controllable thinking effort and agentic coding and tool use.
Those capability descriptions and any associated benchmark results are company-reported claims. They should not be interpreted as proof that Inkling outperforms OpenAI, Anthropic, Google or other models under independently verified, comparable testing.
Inkling-Small
On July 30, 2026, the company announced Inkling-Small, a smaller mixture-of-experts model. Thinking Machines reports that it has:
- 276 billion total parameters;
- 12 billion active parameters;
- Up to a 1-million-token context window;
- Audio and image reasoning capabilities; and
- Availability through Tinker.
These specifications and performance claims come from the company. They describe the model as announced, not as an independently audited comparison or guarantee of production performance.
Research, grants and interactive AI
Thinking Machines’ public work also focuses on human-machine interaction, customizable AI and replicating expert judgment. Its stated mission is to build systems that extend human will and judgment instead of forcing every user into one identical model experience.
The company has published material on human-centered AI, announced interactivity research grants and described research on replicating expert judgment in financial tasks.
What does the NVIDIA partnership change?
On March 10, 2026, Thinking Machines Lab and NVIDIA announced a multi-year strategic partnership involving at least one gigawatt of next-generation NVIDIA Vera Rubin systems. Deployment is targeted for early 2027, according to the company’s announcement.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11NVIDIA also disclosed making a “significant investment” in Thinking Machines Lab, but did not publish the amount in that announcement.
The arrangement matters because frontier-AI development requires more than researchers and algorithms. Companies need large-scale training capacity, inference capacity, predictable access to advanced accelerators and systems-software coordination. A planned gigawatt-scale deployment is therefore evidence of an ambitious infrastructure strategy.
It is not evidence that the company already has a gigawatt of operational computing capacity. Nor does it prove commercial success, revenue, profitability or eventual model superiority. The official announcement describes a deployment targeted for early 2027, not infrastructure already installed and available in August 2026.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How might Thinking Machines make money?
The company’s public products point to several possible or emerging revenue lines:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Usage-based access to the Tinker API;
- Model fine-tuning and post-training;
- Inference and model-serving services;
- Enterprise-specific AI systems;
- Open-weight models that encourage paid Tinker usage; and
- Strategic partnerships with large organizations.
The strongest directly observable commercial product is Tinker. That makes the company relevant to technical teams and researchers evaluating model-customization infrastructure, rather than to ordinary consumers looking for a general chatbot.
Public materials do not establish Thinking Machines Lab’s revenue, customer count, annual recurring revenue, profitability or commercial adoption. A $12 billion post-money valuation reflects what investors agreed to pay in the 2025 transaction; it is not a current independently verified market value and should not be mistaken for evidence of operating performance.
What the valuation does—and does not—tell you
The financing indicates that investors placed substantial value on the company’s team, research ambitions, recruiting power, strategic position and anticipated access to capital and compute.
It does not, by itself, demonstrate:
- Revenue or profits;
- Customer adoption;
- Model superiority;
- Product-market fit;
- A specific founder ownership percentage; or
- Long-term business viability.
It is also misleading to say that Thinking Machines Lab is “worth exactly $10 billion” or “worth exactly $12 billion today.” The $10 billion figure refers approximately to the pre-money valuation reported for the 2025 round, while $12 billion is an approximate post-money figure associated with that transaction. A private startup does not receive a continuously updated public market price in the way a listed company does.
Recommended Free Tools
Can readers use its technology?
Yes. Tinker is the clearest publicly available product connected to Thinking Machines Lab. It offers usage-based access for model customization and fine-tuning, but it is designed for technical users who are comfortable with APIs, model training and post-training workflows.
It may suit researchers and development teams seeking more control over open-weight models. It is less suitable for nontechnical users who simply want a ready-made chat application or for organizations that require a fully managed enterprise application with publicly disclosed production guarantees.
For comparison, the OpenAI API and Anthropic API primarily provide access to their respective hosted-model ecosystems. Hugging Face offers a broader model, dataset and deployment ecosystem. These are alternatives for particular workflows, but they are not identical substitutes for Tinker’s specialized model-customization approach.
What remains unknown?
Even after the company’s product announcements, important financial and corporate details remain private:
- The exact amount raised, beyond descriptions of approximately or close to $2 billion;
- The complete list of investors and each investor’s allocation;
- The exact financing instruments and terms;
- The company’s full capitalization table and dilution;
- Revenue, customer numbers and profitability;
- The extent of commercial adoption for Tinker, Inkling and Inkling-Small;
- Whether the company’s valuation has changed since the 2025 financing; and
- The amount of NVIDIA’s investment.
Those gaps are normal for a private startup, but they limit what can responsibly be inferred from the headline funding number.
Bottom line
Thinking Machines Lab did raise approximately $2 billion in a financing reported in 2025. The $10 billion and $12 billion valuation figures are not necessarily contradictory: approximately $10 billion was the reported pre-money valuation, while approximately $12 billion is the implied post-money valuation after adding the new capital.
The company has since moved beyond its secretive, pre-product phase with Tinker, the Inkling model family, research programs and a planned NVIDIA compute partnership. But the funding round remains evidence of investor confidence—not proof of revenue, profitability, customer traction or future success.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
