{"slug": "ollama-now-supports-jev-style-decision-models", "title": "Ollama now supports Jev-style decision models", "summary": "Ollama 0.35 added support for Jev-style decision models via TypeSafe's Jev API, exposing a new /v1/systemone endpoint that answers a set of named typed questions from a single state input at no additional cost. Three decision models ship with the release: Bespoke Labs' open-source 9B nimble, plus Together AI's experimental 4B tev1 and 0.8B tev1:0.8b. Ollama reported nimble 9B averaged 91ms per decision running locally on an M5 Max, targeting fast tasks such as ticket triage, model routing, and content or safety moderation.", "body_md": "# Ollama now supports Jev-style decision models\n\n## September 29, 2026\n\nOllama now supports decision models, based on [TypeSafe’s Jev API](https://typesafe.ai) for fast, typed decisions:\n\n- No additional costs\n- Lower latency when run locally\n- Three new decision models available today via Ollama\n\nThis new API is available as of Ollama 0.35 by using the new `/v1/systemone` endpoint. Send text as `state` with a set of named questions, and a model running on your machine answers them all in one request. This is great for tasks that require fast decisions, such as ticket triage, model routing, and content or safety moderation.\n\n## Near-instant decisions\n\nDecision models on Ollama are fast, as requests don’t have to travel over a network. Nimble 9B averaged **91ms per decision** in the Pac-Man example below when running locally on an M5 Max. That’s fast enough to make rapid decisions such as playing a game or processing content in real time:\n\n## Available models\n\nThree new decision models are available to run via Ollama:\n\n- [`nimble`](https://ollama.com/library/nimble) : open-source 9B parameter decision model developed by Bespoke Labs\n- [`tev1`](https://ollama.com/library/tev1) : an experimental 4B decision model from Together AI\n- [`tev1:0.8b`](https://ollama.com/library/tev1:0.8b) : an experimental 0.8B decision model from Together AI\n\nMore decision models are coming soon, including models served by Ollama’s cloud.\n\n## Get started\n\nTo get started, first [download](https://ollama.com/download) or upgrade to the latest version of Ollama. Next, download a decision model such as `nimble`:\n\n```\nollama pull nimble\n```\n\nYou can make a request via `curl` or via TypeSafe’s official Python SDK.\n\n**Request**\n\n```\ncurl http://localhost:11434/v1/systemone -d '{\n  \"model\": \"nimble\",\n  \"state\": {\n    \"ticket\": \"I was charged twice. Please refund the extra payment.\"\n  },\n  \"questions\": {\n    \"team\": {\n      \"type\": \"choice\",\n      \"instructions\": \"Which team should handle this ticket?\",\n      \"criteria\": {\n        \"billing\": \"Payments and refunds\",\n        \"technical\": \"Bugs and integrations\",\n        \"other\": \"None of the above\"\n      }\n    },\n    \"refund\": {\n      \"type\": \"noul\",\n      \"instructions\": \"Does the customer explicitly ask for a refund?\"\n    },\n    \"urgency\": {\n      \"type\": \"score\",\n      \"instructions\": \"How urgent is this ticket?\",\n      \"criteria\": [\"Routine\", \"Soon\", \"Urgent\"]\n    }\n  }\n}'\n```\n\n**Response**\n\n```\n{\n  \"model\": \"nimble\",\n  \"answers\": {\n    \"team\": {\n      \"type\": \"choice\",\n      \"choice\": \"billing\",\n      \"probabilities\": {\"billing\": 0.985, \"technical\": 0.012, \"other\": 0.003},\n      \"confidence\": 0.922\n    },\n    \"refund\": {\"type\": \"noul\", \"noul\": 0.997},\n    \"urgency\": {\n      \"type\": \"score\",\n      \"score\": 0.815,\n      \"legend\": {\"0\": \"Routine\", \"1\": \"Soon\", \"2\": \"Urgent\"},\n      \"probabilities\": {\"0\": 0.378, \"1\": 0.429, \"2\": 0.193},\n      \"confidence\": 0.046\n    }\n  },\n  \"usage\": {\"input_tokens\": 841, \"output_tokens\": 4}\n}\n```\n\n**Setup**\n\n```\nuv add typesafe-sdk  # or: pip install typesafe-sdk\nexport TYPESAFE_BASE_URL=http://localhost:11434\nexport TYPESAFE_API_KEY=ollama\nexport TYPESAFE_DEFAULT_MODEL=nimble\n```\n\n**Request**\n\n``` python\nfrom typesafe_sdk import Choice, Noul, Score, TypeSafeClient\n\nticket = \"I was charged twice. Please refund the extra payment.\"\nquestions = {\n    \"team\": Choice(\n        instructions=\"Which team should handle this ticket?\",\n        criteria={\n            \"billing\": \"Payments and refunds\",\n            \"technical\": \"Bugs and integrations\",\n            \"other\": \"None of the above\",\n        },\n    ),\n    \"refund\": Noul(\n        instructions=\"Does the customer explicitly ask for a refund?\",\n    ),\n    \"urgency\": Score(\n        instructions=\"How urgent is this ticket?\",\n        criteria=[\"Routine\", \"Soon\", \"Urgent\"],\n    ),\n}\n\nwith TypeSafeClient(timeout=120) as client:\n    result = client.system_one(\n        state={\"ticket\": ticket},\n        questions=questions,\n    )\n\nprint(result.choices[\"team\"].choice)   # billing\nprint(result.nouls[\"refund\"].noul)     # 0.997\nprint(result.scores[\"urgency\"].score)  # 0.815\n```\n\n## What’s next\n\nThis is the first of many releases to come adding decision model support to Ollama. Future updates will include:\n\n- Faster performance on Apple Silicon powered by MLX\n- More models specializing in different kinds of decision making", "url": "https://wpnews.pro/news/ollama-now-supports-jev-style-decision-models", "canonical_source": "https://ollama.com/blog/ollama-now-supports-jev-style-decision-models", "published_at": "2026-09-30 09:15:33+00:00", "updated_at": "2026-09-30 09:19:38.723444+00:00", "lang": "en", "topics": ["ai-products", "ai-tools", "large-language-models", "ai-agents", "developer-tools"], "entities": ["Ollama", "TypeSafe", "Jev API", "nimble", "Bespoke Labs", "tev1", "Together AI", "M5 Max"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ollama-now-supports-jev-style-decision-models", "markdown": "https://wpnews.pro/news/ollama-now-supports-jev-style-decision-models.md", "text": "https://wpnews.pro/news/ollama-now-supports-jev-style-decision-models.txt", "jsonld": "https://wpnews.pro/news/ollama-now-supports-jev-style-decision-models.jsonld"}}