Organisation
Thinking Machines Lab
Mira Murati's research-and-product company, built from a largely ex-OpenAI team and pitched around adapting AI to specific human needs with more open research than the largest labs.
Thinking Machines Lab is an AI research and product company launched in February 2025 by Mira Murati, formerly OpenAI's chief technology officer, with a founding team drawn heavily from OpenAI including the co-founder John Schulman. It made headlines less for anything it had built than for what its launch signalled about the talent market: five months in, and before shipping any product, it raised a $2 billion seed round at a $12 billion valuation, reported as the largest first funding round on record. It has since released Tinker, a tool for fine-tuning models, and an open-weight model called Inkling, positioning itself around adapting AI to specific human needs and a stated commitment to more open research than the largest labs now practise.
- Category
- Frontier & foundation-model labs
- Founded
- 2025
- HQ
- San Francisco, US
- Funding
- $2bn seed at $12bn valuation (2025), reported the largest first round on record
- Key people
- Mira Murati, John Schulman
Appears alongside
Featured in threads
Tracks
- Models & capabilities 4
- Open weights & ecosystem 3
- Labs & people 2
- Safety & alignment 1
- Ideas & essays 1
- Money & business 1
Thinking Machines Lab sets out staged release framework for open weights
Thinking Machines proposed a staged release process for open-weight models -- inference access, then fine-tuning APIs, then full weights -- to manage dangerous-capability risk.
Open weights & ecosystem · Safety & alignment
Thinking Machines Lab releases Inkling-Small, a distilled open-weight model
Thinking Machines released Inkling-Small, a 276B-parameter (12B active) open-weight MoE model that roughly matches its larger Inkling model despite under a third of the size.
Open weights & ecosystem · Models & capabilities
Thinking Machines Lab releases open-weight model Inkling
The 975-billion-parameter mixture-of-experts model was pitched not as the strongest available but as a base for enterprise fine-tuning through the lab's Tinker platform.
Models & capabilities · Open weights & ecosystem
Thinking Machines Lab previews interaction models
The voice models are built to interrupt and add context mid-conversation rather than wait for a speaker to finish, unlike conventional turn-based voice assistants.
Models & capabilities
Thinking Machines Lab launches Tinker
The former OpenAI CTO's company shipped its first product: a managed fine-tuning API for open-weight models including large mixture-of-experts systems, free during a private beta.
Models & capabilities
Thinking Machines Lab launches Connectionism research blog
The inaugural post identified batch-size-dependent kernels, not floating-point concurrency, as the real cause of non-reproducible LLM outputs, and showed a fix producing identical completions across 1,000 runs.
Ideas & essays
Mira Murati's Thinking Machines Lab raises $2bn seed at $12bn valuation
Thinking Machines Lab, founded by ex-OpenAI CTO Mira Murati, raised the largest seed round on record at $2bn, valuing the five-month-old company at $12bn.
Money & business · Labs & people
Thinking Machines Lab launches
The founding team of roughly 30 researchers, drawn heavily from OpenAI, had no public product and disclosed no funding figure at launch.
Labs & people
Also mentioned in 1 entry
Referenced in passing — Thinking Machines Lab isn't the main subject of these.