TypeSafe AI, the developer of the non-text AI model Jev, has raised $870 million in Series A funding at a $7.5 billion valuation, less than a month after launching its model on September 15, 2026. The funding round was led by Andreessen Horowitz, with participation from Sequoia Capital, existing investor DCVC, and several angel investors.
The company announced the funding on October 9, following rapid adoption of Jev among developers and businesses. TypeSafe claims that approximately one-third of Fortune 500 companies are already using the model, although it has not publicly identified those customers.
Andreessen Horowitz general partner Martin Casado will also join TypeSafe’s board of directors as part of the investment. The company plans to use the funding to develop additional machine-native AI models, improve its infrastructure for software developers, and introduce new enterprise features.
How Jev Differs From Traditional AI Models
Jev uses a transformer architecture similar to those found in large language models (LLMs), but it serves a different purpose. Instead of generating text or code, the model produces probabilities and structured outputs that TypeSafe describes as “calibrated decisions.”
These outputs allow software applications to make decisions directly without first processing natural language responses. TypeSafe claims Jev operates significantly faster and requires substantially fewer tokens than conventional LLMs, making it suitable for automated tasks within software systems.
The company has positioned Jev as an alternative to language-based AI systems for applications that need to process information and execute predefined tasks. Its approach focuses on providing machine-readable decisions rather than conversational responses.
TypeSafe co-founder Diogo Almeida previously explained the distinction in an interview with TechCrunch, noting that existing AI systems have become proficient at human language, while software automation requires a different method of communication.
TypeSafe Plans Further Development Following Rapid Adoption
Jev attracted substantial developer interest immediately after its September 15 release. According to TypeSafe, the model surpassed one million users within days of launching, while Andreessen Horowitz reported that it processed one trillion tokens within its first three days.
The startup also claims that customers have already saved millions of dollars by deploying Jev in production environments. TypeSafe intends to expand its range of machine-native models while developing additional tools and infrastructure for enterprise customers.
TypeSafe was founded in 2024 by Almeida, a former OpenAI researcher, alongside former Meta research engineer Sasha Sheng and engineer and entrepreneur Erik Gafni. The company secured its latest investment just weeks after introducing its first publicly available AI model.
Featured image credits: Magnific.com
For more stories like it, click the +Follow button at the top of this page to follow us.