# Ollama

> Run the laptop build through Ollama; the Gemma listing fits a 16 GB laptop.

Source: https://opendecisio.com/docs/quickstart/ollama/

Ollama packages laptop versions of two bases for its decision route. The Gemma listing is the one for a 16 GB laptop.

 

[Figure: The Gemma listing fits a 16 GB laptop; the Qwen listing needs more. Ollama listings: Gemma q4_k_m, download 7.7 GB; Gemma q4_k_m, in Ollama 8.6 GB; Gemma q8_0, in Ollama 13.6 GB; Qwen, smallest, download 23 GB.]

 

## TL;DR

 

```
# the Gemma 4 12B base, for 16 GB laptops
ollama pull aminroudaki/decisio-gemma
# the Qwen base
ollama pull aminroudaki/decisio
```

 

Run against Ollama 0.35.1 with aminroudaki/decisio-gemma and aminroudaki/decisio, the listing as published, on 2026-10-09; the answers are Ollama's, not the served system's. Its comments are shown above their commands.

 

Then ask a question of the listing you pulled.

  

| Word | Meaning |
| --- | --- |
| listing | a model page on Ollama’s library, with its tags and sizes |
| tag | one build of a listing: `q4_k_m` (also `latest`) and `q8_0` for the Gemma listing |
| decision route | Ollama’s route for typed questions, `/v1/systemone` |

 

Common mistake

Ollama builds its own prompt and applies no calibration, so its numbers are the ones on its model pages, not the ones in this repository. It needs Ollama 0.35.1 or later.

 

## The listings

 

```
# the Gemma 4 12B base, for 16 GB laptops
ollama pull aminroudaki/decisio-gemma
# the Qwen base
ollama pull aminroudaki/decisio
```

 

Run against Ollama 0.35.1 with aminroudaki/decisio-gemma and aminroudaki/decisio, the listing as published, on 2026-10-09; the answers are Ollama's, not the served system's. Its comments are shown above their commands.

 

Requires Ollama 0.35.1 or later. The model pages, [decisio-gemma](https://ollama.com/aminroudaki/decisio-gemma) and [decisio](https://ollama.com/aminroudaki/decisio), package laptop versions of the two bases for Ollama's decision route and list their tags and their sizes.

 

- The Gemma listing is the path for 16 GB laptops: its `q4_k_m` tag (also `latest`) is a 7.7 GB download and takes 8.6 GB in Ollama at the listing's 8,192-token context; its `q8_0` tag takes 13.6 GB (Ollama 0.35.1 on an Apple M5 Pro, [`runs/2026-10-07_ollama-gemma-memory/`](https://github.com/aminry/decisio/tree/v0.11.1/runs/2026-10-07_ollama-gemma-memory)).
- The Qwen listing's smallest tag is a 23 GB download, so it needs a machine with more memory.

 

Ollama builds its own prompt and applies no calibration, so its numbers are the ones on those pages, not the ones in this repository.

 

## Ask a question

 

Send the support-ticket example from [your first request](https://opendecisio.com/docs/first-request/) to Ollama’s decision route, naming the listing. The Gemma listing, the one for 16 GB laptops:

 

```
curl http://localhost:11434/v1/systemone \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "aminroudaki/decisio-gemma",
  "state": "Hi, since this morning none of our 40 staff can log in to the dashboard. We get \"session expired\" right after entering the password. Payroll is due today.",
  "questions": {
    "urgent": {
      "type": "noul",
      "instructions": "Does this ticket need a response within the hour?"
    },
    "category": {
      "type": "choice",
      "instructions": "Which team should handle this ticket?",
      "criteria": {
        "billing": "Invoices, payments, refunds",
        "access": "Login, passwords, permissions",
        "bug": "The product behaves wrongly",
        "other": null
      }
    },
    "impact": {
      "type": "score",
      "instructions": "Rate the business impact using only the reported facts.",
      "criteria": [
        "No function impaired",
        "One user impaired, with a workaround",
        "Many users blocked from a core function",
        "Data loss or legal exposure"
      ]
    }
  }
}'
```

 

Run against Ollama 0.35.1 with aminroudaki/decisio-gemma, the listing as published, on 2026-10-09; the answers are Ollama's, not the served system's.

 

The Qwen listing, which needs a machine with more memory:

 

```
curl http://localhost:11434/v1/systemone \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "aminroudaki/decisio",
  "state": "Hi, since this morning none of our 40 staff can log in to the dashboard. We get \"session expired\" right after entering the password. Payroll is due today.",
  "questions": {
    "urgent": {
      "type": "noul",
      "instructions": "Does this ticket need a response within the hour?"
    },
    "category": {
      "type": "choice",
      "instructions": "Which team should handle this ticket?",
      "criteria": {
        "billing": "Invoices, payments, refunds",
        "access": "Login, passwords, permissions",
        "bug": "The product behaves wrongly",
        "other": null
      }
    },
    "impact": {
      "type": "score",
      "instructions": "Rate the business impact using only the reported facts.",
      "criteria": [
        "No function impaired",
        "One user impaired, with a workaround",
        "Many users blocked from a core function",
        "Data loss or legal exposure"
      ]
    }
  }
}'
```

 

Run against Ollama 0.35.1 with aminroudaki/decisio, the listing as published, on 2026-10-09; the answers are Ollama's, not the served system's.

 

## What next

 

- [Choose your machine](https://opendecisio.com/docs/quickstart/): the other paths.
- [Choosing a base](https://opendecisio.com/docs/bases/): which base is which.
- [Mac with MLX](https://opendecisio.com/docs/quickstart/mac-mlx/): the same laptop, served by decisio itself.