This is a fully managed API to perform inference with pre-trained Small Language Models (SLMs) for various NLP and text processing tasks.

Using SLMs for specific text processing tasks provides three main benefits compared to using LLMs:

- Lower costs: SLMs are much cheaper to run compared to LLMs, making them a cost-effective solution for applications with high inference volumes.

- Improved performance: SLMs are often more accurate and efficient for specific tasks compared to LLMs, which are designed to be general-purpose.

- Lower Latency: SLMs are typically faster to respond compared to LLMs, resulting in lower latency for applications.

This API is meant to provide developers with a simple, efficient and cost-effective way to integrate SLMs into their applications, without the need to self-host them.

All models used in this API were trained by us at Tanaos, are open-source and accessible from the Tanaos Hugging Face page.