ScaleDown models

ScaleDown builds task-specific small language models for context compression, summarization, structured data extraction, and text classification. Its models help developers reduce AI costs and latency by handling focused workloads efficiently, compressing context before it reaches larger models and turning raw text into concise summaries, structured fields, and custom labels.