Send the prompt to the right model before an LLM is called.
What you would have sent to an LLM.
The chosen label, and the score of every candidate.
Live inference runs fastino/GLiNER2.5-Decide on CPU. Each decision is scored independently. Built by Fastino Labs.