Thinking Machines Launches Open Model Inkling to Challenge One-Size-Fits-All AI

Thinking Machines Lab, the AI startup launched by former OpenAI CTO Mira Murati, unveiled its first proprietary model, Inkling, on Wednesday morning. In a deliberate departure from the flagship offerings of OpenAI, Anthropic, and Google, Inkling is open-weight, meaning external developers and companies can download, inspect, and modify the underlying model directly.

Inkling operates as a mixture-of-experts system with 975 billion total parameters. For any specific task, it activates only a fraction of that capacity,roughly 41 billion parameters,a standard architectural choice that keeps massive models faster and more economical to run. The model was trained on 45 trillion tokens spanning text, image, audio, and video, and can reason natively across all four modalities, though its outputs currently remain limited to text, including code, styled artifacts, and structured data.

This release marks Thinking Machines Lab’s first public milestone after a year and a half of largely behind-the-scenes infrastructure work. Some of that effort surfaced earlier in a May research preview of “interaction models,” which were designed to listen, speak, and even interrupt rather than wait passively like typical chatbots. More fundamentally, Inkling tests the startup’s core thesis: that AI systems organizations can adapt for themselves will outperform the one-size-fits-all models the biggest labs currently sell.

Inkling is engineered to deliver calibrated answers, explicitly flagging uncertainty instead of guessing, and allows users to adjust “thinking effort” up or down to prioritize speed. On one benchmark, the company reports Inkling uses a third as many tokens as Nvidia’s Nemotron 3 Ultra to achieve the same coding performance.

The company does not claim Inkling is best-in-class. Its briefing materials state plainly that Inkling is “not the strongest model available today, closed or open.” Instead, the emphasis is on well-rounded performance and customizability.

This raises a central question for the enterprise market Thinking Machines is targeting. For now, the company is marketing Inkling less as a finished product and more as a starting point for organizations to fine-tune through Tinker, its model-customization platform. That means customers, not Thinking Machines, bear responsibility for ensuring their customizations are safe. Fine-tuning, after all, requires serious machine learning talent.

OpenAI, Anthropic, and Google have taken a very different path with ChatGPT, Claude, and Gemini, respectively, building them as general-purpose chatbots first, with agentic features layered on top. Thinking Machines argues that AI trained centrally by one company and frozen in place underperforms AI that organizations shape themselves, because so much expertise is specific to the people who hold it.

That argument is gaining traction. In a blog post published Sunday, Microsoft CEO Satya Nadella, whose company has invested billions in both OpenAI and Anthropic, warned that enterprises using proprietary AI models effectively pay twice: once in subscription costs, and again by handing over business knowledge embedded in their prompts and corrections, which can be absorbed into future model versions.

Hugging Face CEO Clem Delangue made a similar prediction last week. Frontier models, he said, will increasingly be reserved for experimentation and high-value tasks, while most production AI work shifts to private or open-source alternatives,the exact split Thinking Machines is building around.

The clearest evidence for Thinking Machines’ argument came from a recent project with Bridgewater Associates, the world’s largest hedge fund. Researchers from both companies took an existing open-source model and trained it further on Bridgewater’s own financial expertise. The result reportedly scored 84.7% on financial reasoning tests, beating top proprietary AI models while costing roughly a fourteenth as much to run,though those results come from the two companies’ own evaluation, not an independent one.

Either way, Thinking Machines is emphasizing how quickly it got here. OpenAI took roughly five years to bring its tech to market and show revenue, and Anthropic roughly three. Thinking Machines says it did the same in about nine months.

Some will wonder whether Inkling was trained on outputs from competitors’ models, a practice known as distillation that has drawn industry scrutiny. The short answer, per the company’s materials, is partly. Thinking Machines pre-trained Inkling from scratch, but it used other open-weight models, including Moonshot AI’s Kimi K2.5, to help generate some of its early post-training data before large-scale reinforcement learning took over. The next model, the company insists, will use fully self-contained post-training instead.

On the cost side, Thinking Machines has been more guarded. It struck a partnership with Nvidia in March to deploy a gigawatt of Vera Rubin computing capacity and trained Inkling entirely on Nvidia’s GB300 NVL72 systems. The company has not said how it plans to cover those costs, and revenue, by most accounts, has not been a priority. A reported $50 billion fundraising round was said to be coming together last November but had stalled by January; the company has declined to discuss its funding picture since.

A related question is whether Thinking Machines’ spending will ever reach the scale of OpenAI’s or Anthropic’s, or whether its efficiency-driven approach means the economics look different. The company’s bet may be less that it will eventually spend like its larger rivals than that it won’t need to at all. Once weights are public, nothing obligates anyone who downloads them to pay Thinking Machines to run them, unlike the metered access OpenAI and Anthropic sell. It’s Tinker, not the model itself, where the company’s revenue has to come from, via training, fine-tuning, and now a cut of the hosting ecosystem built around it.

Headcount, at least, looks more settled. Thinking Machines now employs roughly 200 people, up from levels reported after a wave of departures earlier this year, including two co-founders who left for OpenAI in January.

Thinking Machines, for its part, does not seem interested in playing up individual moves the way much of the industry does. According to a source inside the company, its culture favors continuity over reliance on any one personality. It makes sense: it’s less of a setback when people change teams if they were never put on a pedestal to begin with. It’s also a remarkable thing for a company to insist on, given how much of its own story is still associated with the name of its now-famous co-founder, whether she planned it or not.