The Fastest-Adopted Model on Vercel: A Game-Changer
The fastest-adopted model on Vercel this year cannot write a word.
TypeSafe Jev, a new AI model, has taken the tech world by storm with its unique capabilities. It returns typed decisions with confidence scores instead of text, eliminating the chance of hallucination due to its predefined answers. Within 24 hours of its launch, it reached an impressive 13% adoption rate among Vercel's paid teams—a record for the AI Gateway.
A Model Beyond Traditional Language
Diogo Almeida, a key figure behind ChatGPT, expresses his disappointment with its current state. In contrast, Jev offers a different approach by refraining from generating text altogether. It takes input in the form of program states and typed questions, providing three types of outputs: choices, scores, or yes-no probabilities.
The Trade-off: Speed and Price
The primary drawback is its inability to produce strings, meaning Jev cannot engage in conversations, drafting, explaining, or coding. However, TypeSafe AI offers a compelling alternative by charging $0.042 per million input tokens ($42 per billion) without metering output. It boasts lightning-fast response times ranging from 70 to 500 milliseconds, significantly outperforming frontier models that take up to 329 seconds for comparable queries.
Reinforcement Learning for Calibrated Decisions
TypeSafe trains Jev using synthetic data generated through reinforcement learning, ensuring precise decision-making. This approach aligns with William Stanley Jevons' economic paradox, where a commodity's falling price increases consumption.
The AI Adoption Curve
Jev's adoption curve is unprecedented, reaching nearly 13% of Vercel's paid teams within hours, surpassing even GPT-5.6 and Fable 5.1. This rapid uptake has left Vercel amazed, as no other model had achieved such a milestone so quickly.
Applications: Streamlining Decision-Making
The model's utility lies in streamlining decision-making processes, especially in tasks that don't require human language. Its speed and efficiency are particularly appealing for startups and tech companies facing growing AI expenses.
Conclusion
Jev represents a significant shift in AI development, prioritizing speed, accuracy, and cost-effectiveness over traditional language capabilities. As the world of AI evolves, models like Jev may become the norm, reshaping the way we interact with technology.