A new artificial intelligence model focused on efficiency and low operating costs was recently introduced. This model is currently the lightest and most cost-effective AI solution, designed for complex, repetitive tasks performed by agents. These tasks include calling external tools, sub-programming tasks, request routing, data extraction, and extensive document analysis.
Significant emphasis is placed on the cost-performance ratio. Priced at just 10 cents per million input tokens, this model appears as a substantially cheaper yet more intelligent alternative compared to some advanced competitor models. Furthermore, it maintains a competitive price even against open models, making it an attractive choice for a wide range of applications.

Got an idea for a website or app?
Let's talk it through — non-binding, with no pressure, and a concrete next step.
Given its focus on low cost and large-scale accuracy, this model is designed to collaborate with larger, more robust models within a complex agent system. A typical example could be customer support, where a more extensive model defines rules and strategies for resolving specific problems, while the lighter model is then used to apply these rules to tens of thousands of individual cases. In other words, while one model plans, the other efficiently processes high-volume operations.
With a view to its use in high-volume scenarios, this artificial intelligence was trained to work effectively with various media types, including text, images, and audio. This makes the model surprisingly flexible, allowing agents to process screenshots, graphs, scanned documents, voice messages, meeting recordings, or video without the need to deploy additional specialized models.
