Thinking Machines Lab is an artificial intelligence research and product development company founded in February 2025 by former OpenAI Chief Technology Officer (CTO) Mira Murati. Headquartered in San Francisco, California, the company focuses on making AI systems more accessible, customizable, and general-purpose, emphasizing human-AI collaboration through multimodal architectures rather than fully autonomous systems. Within a short time after its founding, Thinking Machines Lab drew significant attention and reached a valuation exceeding $10 billion without releasing a final product.

Foundation and Leadership

Mira Murati serves as the founder and Chief Executive Officer (CEO) of Thinking Machines Lab. Previously, she joined OpenAI in 2018, became CTO in 2022, and played a central role in developing projects like ChatGPT, DALL·E, and Codex. After departing OpenAI in late 2023, Murati recruited engineers and researchers from organizations including OpenAI, Character.AI, Meta, and Google DeepMind. The company’s chief scientist is John Schulman, while the CTO is Barret Zoph. Other notable team members include Jonathan Lachman, Alexander Kirillov, Luke Metz, and Andrew Tulloch.

Research Areas and Technological Goals

Thinking Machines Lab focuses on advanced multimodal models designed to enhance human-AI interaction. These systems are intended to function across varying levels of user expertise and integrate into diverse domains. The company prioritizes the development of advanced models that can facilitate transformative applications, particularly in scientific discovery and engineering innovation.

Beyond model intelligence, the company places importance on infrastructure reliability and efficiency. Thinking Machines Lab aims to build robust systems that support long-term research productivity. Multimodality is positioned not only as a means for natural interaction but also as a pathway to better intent recognition and integration with real-world environments.

Product Approach and Safety Principles

The company’s approach to product development is closely integrated with research, with the goal of linking scientific progress to practical applications. Thinking Machines Lab is also active in promoting AI safety. Its policies include preventing model misuse, publishing best practices for safe system design, and openly sharing code, datasets, and model specifications. The company supports both internal development and external research acceleration in the AI safety domain.

Funding and Investors

Despite not having launched a commercial product, Thinking Machines Lab secured $2 billion in early-stage funding. The round was led by Andreessen Horowitz (A16z) and included participation from Sequoia Capital, reflecting investor confidence in the company’s founding team and vision.

Community Engagement

The company commits to regularly publishing its research through technical blogs, academic papers, and open-source code. Emphasizing the collaborative nature of scientific advancement, Thinking Machines Lab aims to contribute transparently to the broader research ecosystem while enriching its own internal research culture.

Structure and Organization

Thinking Machines Lab operates under an interdisciplinary team structure, combining expertise in engineering, research, product design, and experimental sciences. Its development strategy is based on integrating fundamental research with applied development in a unified framework.

Future Vision

The long-term goal of the Thinking Machines Lab is to make AI systems more adaptable, interpretable, and responsive to human needs, transforming how these technologies are integrated into daily life. The company envisions AI models that not only generate knowledge but also support breakthroughs in scientific research, accelerate engineering workflows, and offer solutions across specialized fields.

In the future, the company’s key priorities include creating systems that deliver measurable societal impact and setting new standards in safety and reliability. Regular testing, user feedback loops, and highly adaptable models are central to its development strategy.

Furthermore, Thinking Machines Lab aims to ensure a two-way loop between research and deployment, producing AI systems tested both in controlled environments and in the field by a broad user base. This strategy reflects the company’s intent to build both scientific credibility and commercial viability. Its commitment to human-AI collaboration positions it to pioneer more inclusive and beneficial AI technologies.