Decision Model
A model type distinct from generative large language models that does not produce free-form text but instead makes classification-style selections from a given set of candidate actions or parameters, emitting discrete decisions with low inference cost and predictable behavior.
Core Facts
Timeline (last 90 days)
A Decision Model returns a probability or classification result for content rather than outputting long passages of text, and its structured output can be used directly in if statements in code.
Official data shows that after decision model training, a model's decision accuracy can increase from 30% to 80%.
Decision models train any text or vision LLM via QLoRA and output a confidence score for each candidate option.
Decision models experienced a concentrated explosion between 2024 and 2025, due in part to the popularization of agent orchestration frameworks (such as LangGraph and AutoGen) making 'decision nodes' a system bottleneck, and enterprise deployment needs driving the search for more economical alternatives to full-scale LLMs.
All Facts (4)
A Decision Model returns a probability or classification result for content rather than outputting long passages of text, and its structured output can be used directly in if statements in code.
65%UnverifiedOfficial data shows that after decision model training, a model's decision accuracy can increase from 30% to 80%.
50%UnverifiedDecision models train any text or vision LLM via QLoRA and output a confidence score for each candidate option.
50%UnverifiedDecision models experienced a concentrated explosion between 2024 and 2025, due in part to the popularization of agent orchestration frameworks (such as LangGraph and AutoGen) making 'decision nodes' a system bottleneck, and enterprise deployment needs driving the search for more economical alternatives to full-scale LLMs.
50%