Thinking Machines Laboratory drops its first model

thinking machines lab, Artificial intelligence company Started by exiles from OpenAIreleased its first model, called Inkling. The startup’s new model is open-weight, meaning researchers and startups will be able to download and modify it.

In a Blog postThe company says Inkling has been trained from scratch to understand audio and video input as well as text. She says that while Inkling is not the top model in common benchmarks, it performs well on many tasks, and is capable of advanced reasoning and coding. Like many open-weight models, Inkling is relatively large — 975 billion parameters — and needs to run on a range of specialized chips.

In reference to how artificial intelligence models are growing They are used to build artificial intelligence,The lab also used Inkling to fine-tune and improve itself.

The release could help Thinking Machines cement its place as a legitimate player in the frenetic, big-spending AI race. Open source models have proven popular because they are cheaper to run than closed models, which can usually only be accessed for a fee. Open source models can also be modified more easily for different tasks. The best open-weight models currently come from China, but Thinking Machines says the Inkling offers a level of performance similar to those models.

The release of the Open Weight Model fits into Thinking Machines’ AI vision Another blog post. The company said technology should not be controlled by a few companies and should be decentralized so that more people can build their own models using their own data.

According to a company source who requested anonymity to discuss the development process, researchers discovered a strange phenomenon while training Inkling. Like other models, it usually provides a natural language explanation of its complex logic. Inkling decided to get rid of this in the name of efficiency. “We found that the grammar was above par, which is interesting,” the source says. The company has reused natural language logic to make the models’ decisions more explainable, the person says.

They were thinking machines Founded In February 2025 by several big-name OpenAI executives and researchers, including… Mira Morattiwho served as CTO (and briefly CEO) of OpenAI; John Shulman, a co-founder of OpenAI who played a key role in developing ChatGPT; and Lillian Weng, a former vice president at OpenAI who led work on safety and robotics.

The startup received the largest seed funding round in history, which was valued at US$12 billion. Earlier, the company issued absurditya tool for fine-tuning models, offered a tool that enables Natural voice interactionsHe published research on machine learning.

OpenAI is perhaps the one that started the AI ​​boom with ChatGPTbut maverick-led companies like Thinking Machines and Anthropic Have muscles in space. Anthropic recently filed to go public, valuing the company at more than $1 trillion. Her model Claude has proven popular with many companies, especially for his programming skills.