Workflows/Run a private reasoning assistant with gpt-oss · 120B

Text / Intermediate

Run a private reasoning assistant with gpt-oss · 120B

Install Ollama, pull the gpt-oss:120b model, start it, and send a reasoning-heavy prompt to verify the model responds correctly.

EstimatedSource checked 9/24/2026

LOCALRENTED GPU

Before you begin

Difficulty: Intermediate
Software: Ollama
Hardware: For the publisher's 65.0 GB Ollama text package at a short 4K context, we estimate at least 88 GB GPU memory or 112 GB Apple unified memory. For a more comfortable starting point, use 120 GB GPU memory or 168 GB unified memory. These are capacity estimates, not speed tests; longer context and other apps need additional headroom.

TEXT / OLLAMASOURCE-LINKED SETUP

Run a private reasoning assistant with gpt-oss · 120B, step by step.

Install Ollama, pull the gpt-oss:120b model, start it, and send a reasoning-heavy prompt to verify the model responds correctly.

Choose a package Run its commands Check the result

BEFORE YOU BEGIN

The documented package

This setup uses the package documented for this task. Review its source and supported platforms before starting.

Save your machine in My Hardware to get an automatic starting choice. You can always choose any package yourself.

CURRENT SETUP

gpt-oss · 120B · official

For the publisher's 65.0 GB Ollama text package at a short 4K context, we estimate at least 88 GB GPU memory or 112 GB Apple unified memory. For a more comfortable starting point, use 120 GB GPU memory or 168 GB unified memory. These are capacity estimates, not speed tests; longer context and other apps need additional headroom.

View weight source

THE WORKFLOW

Set up gpt-oss · 120B on your machine

Pick your operating system. Every command below is for the selected package and runtime.

Install Ollama for this operating system before running the model command.

Install Ollama

Download and install Ollama for Windows, macOS, or Linux from the official download page. Follow the installer's prompts; no terminal command is required for this step.

Pull the 120B model

In a terminal, download the gpt-oss 120B package. This is a large file (~65 GB), so allow time and disk space.

ollama pull gpt-oss:120b

Start the model and open an interactive session

Launch the model in interactive mode. This loads the model and gives you a prompt where you can type messages.

ollama run gpt-oss:120b

Set reasoning level to high

At the interactive prompt, type the system-style instruction to request maximum reasoning effort for the next response. Ollama passes this as part of the conversation context.

Reasoning: high

Send a reasoning task

Copy this prompt exactly to test the model's ability to produce a detailed, step-by-step reasoning response: Type this as the local chat prompt: “A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? Explain your reasoning step by step.”

Check the model's response

Verify that the model outputs a coherent explanation that works through the algebra and concludes the ball costs $0.05. A correct response will show intermediate reasoning, not just the final number.

SUCCESS CHECK · gpt-oss · 120B

Check a known reasoning problem

Try a short multi-step problem with a known result. Verify the final answer and key steps independently before increasing context.

This is a source-linked setup, not a YouRunAI hardware test. Confirm your exact runtime version, package, and output before relying on it.

Back to setup steps

When it doesn’t go to plan

The model download fails or stalls

Ensure you have at least 70 GB free disk space and a stable internet connection. Run 'ollama pull gpt-oss:120b' again; Ollama resumes partial downloads.

Ollama reports insufficient memory when loading gpt-oss:120b

The 120B model requires a GPU with about 80 GB VRAM (e.g., H100 or MI300X). If your hardware is smaller, switch to the 20B variant by running 'ollama run gpt-oss:20b' instead; the rest of the steps apply the same way.

The model responds but ignores the reasoning instruction

Start a fresh session with 'ollama run gpt-oss:120b' and include 'Reasoning: high' as the first line of your prompt, followed by a newline and then your question.

The response is cut off or the session hangs

Press Ctrl+C to interrupt, then restart with 'ollama run gpt-oss:120b'. If the issue persists, reduce the reasoning level by sending 'Reasoning: medium' before your prompt.

REFERENCE LIBRARY

Sources and files

Original instructions, model files, and compatibility notes behind this setup.

4 SOURCES

01Publisher model card and weightshuggingface.co 02Publisher run instructionsollama.com 03Official Ollama downloadollama.com 04Ollama CLI referencedocs.ollama.com

The model behind this workflow

gpt-oss · 120B

MY HARDWARE

Will it run on your machine?

Save your machine to see a personalized rating and its reasoning.

Add my hardware

Keep exploring

Text

Beginner

Set up a private AI chat assistant

Install a local runtime, run Qwen3.5 9B, confirm responses, and know when to choose the smaller 4B package.

Build a private local coding assistant

Connect an open coding-capable model in Ollama to Cline, run a small repository task, and review the result locally.

SOFTWAREOllama + Cline

Generate images on your own GPU

Use the official FLUX.2 Klein 4B ComfyUI template with exact model files, a first prompt, and an output check.

SOFTWAREComfyUI

LOCALRENTED GPU

Verified source