Background task-man for akro.

tools thinking

ollama run akropora/neuron

curl http://localhost:11434/api/chat \
  -d '{
    "model": "akropora/neuron",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='akropora/neuron',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'akropora/neuron',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code ollama launch claude --model akropora/neuron

OpenCode ollama launch opencode --model akropora/neuron

Hermes Agent ollama launch hermes --model akropora/neuron

OpenClaw ollama launch openclaw --model akropora/neuron

Models

View all →

Name

1 model

Size / Usage

Context

Input

neuron:latest

523MB · 40K context window · Text · 2 months ago

neuron:latest

523MB

40K

Text

Readme

Neuron

Neuron is a tiny utility model built for Akro, a local AI environment focused on small models, memory, retrieval, and modular tools.

Neuron handles fast background jobs such as:

chat naming
lightweight classification
candidate selection
structured JSON utility tasks

It is designed to be small, fast, and predictable rather than conversational.

Pull

ollama pull akropora/neuron:latest

Use with Akro

Akro expects the local name:

neuron:latest

Create that alias with:

ollama cp akropora/neuron:latest neuron:latest

Neuron is based on Qwen3 0.6B and is tuned for short, structured internal tasks.