Model overview
What this model is for
Decision-2.0-Kai-0.6B is a text model designed for typed decisions rather than open-ended prose generation. It handles caller-defined choice, yes/no, and score questions by returning probability distributions, making it suitable for routing, classification, ranking, and other applications that need structured decision outputs. The model is part of vLLM Semantic Router’s Decision 2.0 family and is identified as an Apache-2.0 release, so developers should review the license and the model documentation before incorporating it into a product. Its documented modality is text, and its primary listed use case is typed decisions. The available source does not disclose a context-window size, inference throughput, pricing, detailed evaluation results, or additional capabilities. It should therefore be assessed with application-specific tests for calibration, accuracy, and output integration before deployment. The supplied Hugging Face page is the official reference, but the retrieved page content does not expose a model-card update date or further technical documentation, so source freshness beyond the current page retrieval is not disclosed. The model is listed at approximately 0.6 billion parameters.