According to CoinMeta, Thinking Machines Lab, founded by CTO Mira Murati before OpenAI, has released its first general-purpose model, Inkling. The model has a total of 975 billion parameters, with 41 billion parameters activated each time, and it supports text, images, and audio, with a maximum context length of 1 million token. This model was trained from scratch, and its architecture is mainly based on DeepSeek-V3; the complete weights have been made available on Hugging Face. Meanwhile, Thinking Machines Lab also released a model, Inkling-small, which has less than 30% of the parameters of Inkling. Its performance is only 1 point lower, and it also supports text, image, and audio input, using the Apache 2.0 license.