Thinking Machines Lab
Builds multimodal foundation models and fine-tuning infrastructure.
Available information varies by company and source.
Profile record updated:
Company facts
- Official name
- Thinking Machines Lab, Inc.
- Entity type
- COMPANY
- Founded
- 2025
- Headquarters
- United States
- Company size
- 50–200
- Market role
- B2B SaaS Provider
- Official website
- thinkingmachines.ai
What Thinking Machines Lab does
Thinking Machines Lab creates value by developing proprietary multimodal AI models and making them usable through a managed developer platform. Open-weight releases widen adoption among researchers and engineering teams, while the company captures revenue from the operational layer: API access, token usage, model fine-tuning, evaluation workflows and related platform services. This structure blends research-led distribution with software monetisation.
Category differentiation
This company is a foundation-model lab and developer platform, not a media, advertising or marketing technology vendor. It should not be confused with historical Thinking Machines Corporation or with generic AI consultancy firms.
Strategic context
AI-supported assessment from the existing company research; distinguish interpretation from sourced facts.
Thinking Machines Lab is a private United States AI research and product company founded in 2025. It develops multimodal foundation models and a developer platform that allows organisations to fine-tune, evaluate and integrate those models into their own products and workflows. Its product set includes the open-weight model Inkling, the real-time multimodal Interaction initiative and the Tinker platform for token generation, experimentation and model operations. The company serves AI researchers, machine learning engineers, developer teams and enterprise product builders. Its commercial model combines open-weight model releases to drive adoption with paid platform usage on Tinker, where customers consume inference, fine-tuning and evaluation tooling. The business is positioned around transparency, customisability and developer control rather than a closed hosted-model-only approach.
Company news briefing
Briefing updated:
Thinking Machines Lab, founded by Mira Murati, is reportedly in talks with Accel to lead a $1 billion funding round at a valuation of at least $40 billion, supported by an annual revenue run rate exceeding $100 million from its Tinker platform. Meanwhile, the startup continues to collaborate within Nvidia's Nemotron Coalition to advance open-weight models aimed at accelerating industrial and enterprise AI adoption.
Business model & monetisation
The company monetises through a hybrid open-weight and platform model. Inkling and related model assets drive adoption through research access and downstream integration, while paid revenue is generated through Tinker via API and platform access, token generation, compute consumption, fine-tuning workflows and evaluation tooling. The dominant pricing logic is pay-per-use infrastructure consumption, potentially supplemented by enterprise platform agreements.
- Tinker platform usage
- Pay-per-Use
- Enterprise platform access and support
- Software Subscription
- Model fine-tuning and evaluation workflows
- Pay-per-Use
- Open-weight model distribution
Products & capabilities
No products with linked sources are available in this view.
Products & market categories
Technology
Recent recorded signals
Dates refer to the source publication. Older entries are historical context, not evidence of a new event.
Barret Zoph Joins Google as VP of Research
Large Language Models (LLM) & AI · Recorded impact score: 2/5
Barret Zoph, co‑founder of AI startup Thinking Machines Lab and a former OpenAI employee, has taken a role as vice president of research at Google. Zoph previously co‑founded Thinking Machines with Mira Murati after leaving OpenAI in October 2024, briefly returned to OpenAI in January 2026 to lead AI enterprise sales, and departed in June 2026 after a five‑month stint. Tech reporting notes Zoph was fired from Thinking Machines earlier this year. Google said it expects Zoph to contribute reinforcement learning and post‑training expertise to its Gemini efforts. The move is one of several high‑profile executive shifts in the AI industry this year.
- Barret Zoph co‑founded Thinking Machines Lab and later rejoined OpenAI.
- Zoph spent five months at OpenAI in 2026 heading AI enterprise sales and left in June 2026.
Putting Task Expertise into RL Achieves State-of-the-Art Performance on Text-to-SQL
Recorded impact score: 3.5/5
Thinking Machines Lab announced two new updates: a research breakthrough in RL for text-to-SQL and new safety research grants.
Thinking Machines releases open-weight model Inkling
Large Language Models (LLM) & AI · Recorded impact score: 3/5
Thinking Machines released Inkling, an open-weights, general-purpose multimodal Mixture-of-Experts foundation model: a 66-layer decoder-only transformer with a sparse MoE backbone (≈975B total parameters, ≈41B active per task). It natively accepts text, images and audio and produces UTF-8 text outputs. Trained from scratch on roughly 45 trillion tokens assembled from public, third-party and synthetic sources on NVIDIA GB300 NVL72 systems under a strategic NVIDIA partnership, Inkling ships with checkpointed support for up to a 1,000,000-token context (Tinker API exposes 256K) and includes a smaller preview (Inkling‑Small, ≈276B/12B). Published under Apache‑2.0 with day‑0 ecosystem and inference/hosting support (Hugging Face, vLLM, SGLang, TokenSpeed, Modal, Databricks, Baseten, NVIDIA optimizations and community quantization), it is positioned for enterprise fine‑tuning via Tinker, evaluated against benchmarks (2026-07-14) and accompanied by safety guidance.
- Architecture: 66-layer decoder-only transformer with a sparse MoE backbone — ≈975B total parameters, ≈41B active per task; accepts text, image and audio inputs and outputs UTF-8 text.
- Training: trained from scratch on roughly 45 trillion tokens drawn from public, third-party and synthetic/augmented datasets; training ran on NVIDIA GB300 NVL72 systems under a strategic NVIDIA partnership.
AI Demands New Interaction Models for Designers
Interaction Models / Conversational UX · Recorded impact score: 2/5
The article argues that recent multimodal AI models are changing the fundamental grammar of software interaction, shifting interfaces from task-driven, turn-taking flows to continuous, intent-driven exchanges. It highlights a research preview from Thinking Machines Lab that demonstrates real-time multimodal 'interaction models' which can listen, see, and respond across audio, video, and text while building interfaces (a generative UI) on the fly. The author outlines design implications — rebuilding mental models, new entry/navigation conventions, deliberate intervention points, and routing judgment to humans — particularly for enterprise contexts that require auditability and accountability.
- Thinking Machines Lab released a research preview called Interaction Models demonstrating real-time multimodal exchange (audio, video, and text) and on-the-spot interface generation.
- The article’s author, Arin Bhowmick, is Chief Design Officer at SAP.
Explore company relationships
Questions about Thinking Machines Lab
What is Thinking Machines Lab?
Thinking Machines Lab is a private AI research and product company that builds multimodal foundation models and a developer platform for fine-tuning, evaluation and integration.
Who uses Thinking Machines Lab?
Its users are AI researchers, machine learning engineers, developers and enterprise product teams that need customisable foundation models and model tooling.
How does Thinking Machines Lab make money?
It monetises through platform usage on Tinker, including token consumption, fine-tuning, evaluation and enterprise platform access rather than charging solely for model access.
Sources & coverage
This profile uses public, official and technically observable information. Missing information does not prove that a product or relationship does not exist. The list below does not imply that every profile statement has been verified.
15 publicly documented primary sources and citations linked across the market graph.
Continue your research on Thinking Machines Lab
Explorer includes additional company details, a Watchlist for up to 25 companies and your personal Strategic Intelligence Agent. It monitors your market daily and delivers tailored briefings with clear strategic context whenever relevant news occurs.
Free, with no time limit.