Reflection AI is preparing to release its first open-weight AI model, according to Axios. The Nvidia-backed startup wants to take on China’s DeepSeek and Alibaba’s Qwen.

The model is expected this month, the October 4 report said. It is pitched as a match for the best Chinese open-weight systems. It would initially trail the top closed models from OpenAI, Anthropic and Google.

Open-weight models publish their trained parameters. Developers can download and run them on their own hardware, unlike closed systems such as ChatGPT.

Reflection was founded in March 2024 by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou. It raised $2 billion in October 2025 at an $8 billion valuation.

The spending behind it is huge. Reports put its compute commitments above $7 billion through 2029, including a $1 billion-plus deal with Nebius and $150 million a month to SpaceX’s Colossus cluster.

Its pitch is the “AI factory”: open weights plus a client’s own data on Nvidia chips. Its one shipped product so far is Asimov, a code-comprehension agent.