Sunday, October 18, 2026
Herald of AI

AIGenerated AI news, written by AI and explained simply. Models, technology, culture and the people behind it.

New models

Non-text AI model Jev lands $7.5B valuation weeks after launch

TypeSafe AI's model skips text entirely and goes straight to decisions, attracting top investors.

Elegant desk lamp illuminating a closed laptop on a desk, creating a warm and cozy atmosphere.
Elegant desk lamp illuminating a closed laptop on a desk, creating a warm and cozy atmosphere. File picturePhoto: Shan Nir / Pexels

Just weeks after launching its unusual AI model, [TypeSafe AI] has raised $870 million at a $7.5 billion valuation. The round was led by Andreessen Horowitz, with participation from Sequoia and existing backer DCVC 1.

A Model That Doesn't Write

The model behind this funding round is called [Jev], and it does something most AI systems don't: it doesn't produce text at all. Instead of writing sentences or answering questions, Jev outputs what the company calls "calibrated decisions" — probabilities that tell a computer what to do next 3.

Jev is built on a transformer architecture, the same kind of system behind most modern AI language models. But where a standard large language model generates the next word in a sentence, Jev evaluates inputs and returns a decision-weighted answer 2.

Think of it this way: a language model might write you an email about a shipping delay. Jev would simply calculate the best route for that shipment and output the decision — no words needed 3.

Quick Adoption by Big Companies

According to TypeSafe AI, one-third of Fortune 500 companies are already using Jev 1. That is an unusually fast climb for a new AI product, especially one that doesn't look like the chatbots and writing tools people have grown used to.

The company says Jev runs significantly faster and uses far fewer tokens — the units of text that AI models process — than traditional large language models 2. That efficiency matters a lot when you are running automated systems at scale.

Built by veterans

TypeSafe AI was co-founded in 2024 by three people with serious resumes in AI research: [Diogo Almeida], formerly a researcher at OpenAI; [Sasha Sheng], a former Meta research engineer; and [Erik Gafni], an engineer and entrepreneur 3.

Almeida explained the reasoning behind building a non-text model: "We have been super good at human language for four years, but it's not useful for automation because computers speak a different language" 1.

Jev is available through an API and a web interface, though the company has not shared specific pricing details 2. It is designed primarily for task automation rather than creative or conversational work.

Key facts

  • TypeSafe AI raised $870 million at a $7.5 billion valuation
  • Jev outputs probabilities and decisions, not text
  • One-third of Fortune 500 companies reportedly use Jev
  • Founded in 2024 by ex-OpenAI, ex-Meta researchers
  • Round led by Andreessen Horowitz with Sequoia and DCVC

Sources

  1. The maker of non-text AI model Jev valued at $7.5B just weeks after launch | TechCrunch — techcrunch.com
  2. TypeSafe AI Raises $870M at $7.5B Valuation for Non-Text Model Jev — aitoolsoasis.com
  3. TypeSafe AI Jev Hits $7.5B Valuation 2026 - shattered.io — shattered.io
  4. TypeSafe Raises $870M at $7.5B for Non-Text AI Model Jev, October 2026 ... — kurums.com
  5. TypeSafe's Non-Text AI Model Jev Valued at $7.5B Post-Launch — renascence.io