Reflection AI, a startup founded in 2024 by former DeepMind researchers, has raised $2 billion at an $8 billion valuation. Initially focused on autonomous coding agents, the company now aims to serve as an open-source alternative to closed labs like OpenAI and Anthropic, while also competing with Chinese AI firms such as DeepSeek.
The startup has recruited talent from DeepMind and OpenAI, built an advanced AI training stack, and secured a compute cluster. Laskin said the company plans to release a frontier language model next year trained on “tens of trillions of tokens.” He added, “We built something once thought possible only inside the world’s top labs: a large-scale LLM and reinforcement learning platform capable of training massive Mixture-of-Experts (MoEs) models at frontier scale.”
Reflection AI’s approach balances open access with commercial strategy, releasing model weights for public use while keeping datasets and training pipelines proprietary. Revenue will come from enterprises and governments building products or sovereign AI systems using the models. Laskin noted, “Once you get into that territory where you’re a large enterprise, by default you want an open model. You want something you will have ownership over. You can run it on your infrastructure. You can control its costs. You can customize it for various workloads.” Investors include Nvidia, Sequoia, CRV, Citi, and Eric Schmidt, among others.