{"slug": "kev-decision-model-in-llama", "title": "Kev Decision Model in Llama", "summary": "A blogger tested Kev, an open-weights decision model distributed as Kev-4B-Q4_K_M.gguf (a 3 GB file based on Qwen 3.5 4B) and served through Llama, which returns JSON answers with probabilities instead of generated text. Asked to classify \"the sky is azure\" against blue, green, red, black and yellow, Kev returned blue at 0.96 and green at 0.02, with all other colors at 0.006 or below; for \"it is night time in the countryside\" it returned black at 0.82, blue at 0.08 and red at 0.06, responding instantly in both tests.", "body_md": "There has been some excitement over **decision models** in the AI world lately. A decision model is like an LLM but it does not generate text in its output. It accepts a prompt, but returns **JSON output** and a list of **probabilities**—so basically you pass it a question and it responds with the answer (or likelihood of the answers).\n\nDecision models are on the `systemone` end point. I downloaded `Kev-4B-Q4_K_M.gguf`, which is an open-weights version of Jev called Kev. It is based on Qwen 3.5 4B, and is only 3 GB. I then set up **Llama** to use this model.\n\nWith `curl` commands, I can send questions and instructions to the model. For example, I used the **question** \"what color is the sky?\" and offered the **criteria** blue, green, red, black, yellow.\n\nWhen I used the prompt \"the sky is azure\", **Kev responded** with \"blue\" and \"green\" with probabilities 0.96 and 0.02; the other colors were all 0.006 or below. So it understood \"azure\" means blue and green. Then I tried \"it is night time in the countryside.\" Kev responded with the probabilities black (0.82) and blue (0.08) and red (0.06). The model linked \"night time\" with a dark sky, possibly with blue or red (due to the sunset). It responded instantly in both tests; Kev is a promising way to classify or decide based on prompts.", "url": "https://wpnews.pro/news/kev-decision-model-in-llama", "canonical_source": "https://www.dotnetperls.com/2026_11_18_kev-decision-model-llama", "published_at": "2026-11-18 07:00:00+00:00", "updated_at": "2026-11-18 17:41:59.470182+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "large-language-models", "ai-tools"], "entities": ["Kev", "Kev-4B-Q4_K_M.gguf", "Qwen 3.5 4B", "Llama", "Jev"], "also_reported_by": [], "takeaway": "A 3 GB Qwen 3.5 4B-based decision model served in Llama returned calibrated JSON probabilities for constrained-choice prompts, scoring \"the sky is azure\" as blue 0.96 and green 0.02 and \"night time in the countryside\" as black 0.82, blue 0.08 and red 0.06.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/kev-decision-model-in-llama", "markdown": "https://wpnews.pro/news/kev-decision-model-in-llama.md", "text": "https://wpnews.pro/news/kev-decision-model-in-llama.txt", "jsonld": "https://wpnews.pro/news/kev-decision-model-in-llama.jsonld"}}