The engineering agent that lives on your machine and works like it's the last one you'll ever need. Local. Private. Sharp as onyx.

tools

ollama run Natuworkguy/flash-onyx-1

curl http://localhost:11434/api/chat \
  -d '{
    "model": "Natuworkguy/flash-onyx-1",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='Natuworkguy/flash-onyx-1',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'Natuworkguy/flash-onyx-1',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 23 hours ago

23 hours ago

046c1dc58d7a · 4.9GB ·

model

archllama

parameters8.03B

quantizationQ4_K_M

4.9GB

template

{{- if or .System .Tools }}<|start_header_id|>system<|end_header_id|> {{- if .System }} {{ .System }

1.5kB

license

LLAMA 3.1 COMMUNITY LICENSE AGREEMENT Llama 3.1 Version Release Date: July 23, 2024 “Agreement”

12kB

system

You are Flash Onyx 1, the flagship model of FLASH (Fast Local Agent SHell). Not a chatbot, not an as

3.3kB

params

{ "min_p": 0.05, "num_ctx": 16384, "num_predict": 2048, "repeat_last_n": 128, "r

221B

Readme

Flash Onyx 1

A local-first engineering agent that thinks like your sharpest teammate and talks like they’re texting you. No cloud, no telemetry, no leash. Onyx reads your code, traces bugs to the root, and hands back the fix in two lines flat. Everything runs on your machine, so nothing you touch ever leaves it.

Pull it

ollama pull Natuworkguy/flash-onyx-1

Run it

ollama run Natuworkguy/flash-onyx-1

Or wire it into the FLASH CLI:

MODEL=Natuworkguy/flash-onyx-1

Built on llama3.1 8B, tuned for tool use and a terse, decisive voice
16K context, function-calling ready
Fast, private, ruthlessly precise