TypeSafe AI, maker of the Jev model, raises $870 million
AI startup TypeSafe AI has raised $870 million at a $7.5 billion valuation. Its Jev model, which is not a language model, became popular rapidly after its mid-September release.
TypeSafe AI, a US artificial intelligence startup behind a new type of model called Jev, announced on October 9 that it has raised $870 million. The funding round values the company at $7.5 billion, just weeks after the model was released. It shows how quickly demand is growing for AI built to automate tasks rather than to generate text.
The round was led by venture capital firm Andreessen Horowitz. Sequoia also took part, as did existing investor DCVC.
What Jev is
Jev was released on September 15 and became popular almost immediately. It is built on a transformer architecture, which large language models (LLMs) also use, but it is not an LLM itself. Instead of producing text, it outputs probabilities, which the company calls "calibrated decisions."
TypeSafe claims Jev runs significantly faster than LLMs and uses far fewer tokens, the units of data in which models process information. The company says its approach is particularly well suited to automating tasks rather than generating text or code.
Enterprise adoption
According to the startup, a third of Fortune 500 companies, the largest US firms by revenue, are already using the model.
TypeSafe co-founder Diogo Almeida, previously a researcher at OpenAI, told TechCrunch last month that the industry has been very good at human language for four years, but that this is not useful for automation because computers speak a different language.
TypeSafe was founded in 2024 by Almeida together with Sasha Sheng, a former Meta research engineer, and Erik Gafni, an engineer and entrepreneur.
