Mindket
English
English简体中文
Sign inSign up
StoreMCPSkillsCalculatorLearnPriceBlog
Discover
Research311Writing59Business95Legal13Finance & accounting33Marketing & SEO175Developer1,420Productivity390Education81Companions14
Help and support
Language
English简体中文

gpt-oss-20b-GGUF by ggml-org

Made by ggml-orgKitDeveloper
Newversion 1 · updated
Like

Overview

Generates text on your own computer from a GGUF copy of OpenAI's gpt-oss-20b, packaged by ggml-org. For people who want to run an open 20B model locally in the Llama app or from a terminal.

gpt-oss-20b-GGUF by ggml-org is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page, a zip of its files. Nothing in it runs by itself.

At a glance

You need
  • The Llama app, or the llama command in a terminal
  • Python, only if you copy the files yourself
  • Room for roughly 10.5 GB of weights, which is the estimate for a 4-bit copy
Runs in
  • The Llama app
  • A terminal, with the llama serve command
  • Windows, macOS or Linux with Python, for copying the files with Hugging Face's own steps
Start with
llama serve -hf ggml-org/gpt-oss-20b-GGUF

Platform

WindowsOperating system
macOSOperating system
LinuxOperating system

About

What it is

gpt-oss-20b-GGUF is a GGUF copy of OpenAI's gpt-oss-20b text generation model, packaged by ggml-org. Hugging Face lists 20.9 billion parameters. The card says the files were converted automatically with ggml-org's own convert tool, and it gives a short way to run them.

Who it suits

People who want to run an open 20B text model on their own computer instead of calling a hosted service. The files carry the Apache 2.0 license.

IndustryWho uses itHow they use itExample
SoftwareDevelopersRun a local model server for apps and testsStarting the serve command against this repository
ResearchHobbyistsTry an open model without a hosted APIChatting with the model in the Llama app
EducationStudentsLearn how local models are packaged and runComparing a GGUF copy with the original model's page

Anyone after benchmarks, tuning tips or the unconverted weights can skip it, because the card is short and points to the source model for those.

Where to use it

  • The Llama app
  • A terminal, with the llama serve command
  • Windows, macOS or Linux with Python, for copying the files with Hugging Face's own steps
Use caseWhoWhat they get back
Local chatEveryday usersGenerated replies on your own machine
Local serverDevelopersA model endpoint started from the terminal
Offline trialsResearchersAn open model you can test without an API key

How to set it up

  • The Llama app, or the llama command in a terminal
  • Python, only if you copy the files yourself
  • Room for roughly 10.5 GB of weights, which is the estimate for a 4-bit copy
  1. Install the Llama app, or make sure the llama command works in your terminal.
  2. Run the serve command below, which points at this Hugging Face repository.
  3. Or copy the files yourself with Hugging Face's two commands in the second block.
llama serve -hf ggml-org/gpt-oss-20b-GGUF
pip install -U huggingface_hub
hf download ggml-org/gpt-oss-20b-GGUF

How to use it

Open the model in the Llama app, or start it from a terminal with the serve command above. The card gives no other settings or prompts, so you chat with it the way you would with any local text model.

Example

You run the serve command with this repository and then send the running model a question. It answers with generated text, as a gpt-oss-20b model does.

For newcomers and experts

GGUF is a packaged file format for running models on your own computer. A good first step is to run the serve command once and see the model load.

The card names two source models, OpenAI's gpt-oss-20b and a RedHatAI eagle3 speculator for it. It does not say which quantization the files use, so check the repository's file list before you plan around the size.

  1. Pick the Llama app or a terminal
  2. Run llama serve with this repository
  3. The GGUF model loads
  4. Send a prompt
  5. Read the generated reply

Good to know

  • License: Apache 2.0
  • Cost to run: none named
  • Data it reaches: the download fetches files from Hugging Face, and the source names nothing else
  • The card is brief and lists no benchmarks, quantization types or sampling settings

README and files

README

gpt-oss-20b-GGUF by ggml-org

Generates text on your own computer from a GGUF copy of OpenAI's gpt-oss-20b, packaged by ggml-org. For people who want to run an open 20B model locally in the Llama app or from a terminal.

Made by ggml-org on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: gpt-oss-20b-GGUF is a GGUF copy of OpenAI's gpt-oss-20b text generation model, packaged by ggml-org. Hugging Face lists 20.9 billion parameters. The card says the files were converted automatically with ggml-org's own convert tool, and it gives a short way to run them.

Who it suits: People who want to run an open 20B text model on their own computer instead of calling a hosted service. The files carry the Apache 2.0 license.

Industry | Who uses it | How they use it | Example
Software | Developers | Run a local model server for apps and tests | Starting the serve command against this repository
Research | Hobbyists | Try an open model without a hosted API | Chatting with the model in the Llama app
Education | Students | Learn how local models are packaged and run | Comparing a GGUF copy with the original model's page

Anyone after benchmarks, tuning tips or the unconverted weights can skip it, because the card is short and points to the source model for those.

Where to use it:

  • The Llama app
  • A terminal, with the llama serve command
  • Windows, macOS or Linux with Python, for copying the files with Hugging Face's own steps
Use case | Who | What they get back
Local chat | Everyday users | Generated replies on your own machine
Local server | Developers | A model endpoint started from the terminal
Offline trials | Researchers | An open model you can test without an API key

How to set it up:

  • The Llama app, or the llama command in a terminal
  • Python, only if you copy the files yourself
  • Room for roughly 10.5 GB of weights, which is the estimate for a 4-bit copy
  1. Install the Llama app, or make sure the llama command works in your terminal.
  2. Run the serve command below, which points at this Hugging Face repository.
  3. Or copy the files yourself with Hugging Face's two commands in the second block.
llama serve -hf ggml-org/gpt-oss-20b-GGUF
pip install -U huggingface_hub
hf download ggml-org/gpt-oss-20b-GGUF

How to use it: Open the model in the Llama app, or start it from a terminal with the serve command above. The card gives no other settings or prompts, so you chat with it the way you would with any local text model.

Example: You run the serve command with this repository and then send the running model a question. It answers with generated text, as a gpt-oss-20b model does.

For newcomers and experts: GGUF is a packaged file format for running models on your own computer. A good first step is to run the serve command once and see the model load.

The card names two source models, OpenAI's gpt-oss-20b and a RedHatAI eagle3 speculator for it. It does not say which quantization the files use, so check the repository's file list before you plan around the size.

Pick the Llama app or a terminal
Run llama serve with this repository
The GGUF model loads
Send a prompt
Read the generated reply

Good to know:

  • License: Apache 2.0
  • Cost to run: none named
  • Data it reaches: the download fetches files from Hugging Face, and the source names nothing else
  • The card is brief and lists no benchmarks, quantization types or sampling settings

On Hugging Face, 13 November 2026

  • Task: text generation
  • Parameters: 20.9B
  • Weights: about 10.5 GB at 4-bit (estimate)
  • Formats: GGUF
  • Downloads in the last 30 days: 371,075
  • Likes: 200
  • License: apache-2.0
  • Based on: openai/gpt-oss-20b
  • Tags: gguf, quantized, text-generation, conversational

Links

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download ggml-org/gpt-oss-20b-GGUF

Then follow the usage in its model card. Nothing in this kit runs by itself.

Files
README.md  4 KB

1 file, 4 KB, downloaded as a zip.

What’s new

  • Version 1 · 2026-11-13

Compare with similar

Listed Hugging Face models in the same category that share the most tags with this one.

ProductDownloads, last 30 daysLikesParametersLicenseLast change
gpt-oss-20b-GGUF by ggml-org This page371,07520020.9Bapache-2.0
Qwen3.8-35B-A3B-Distill-GGUFShares quantized · gguf · text-generation1,337,09727735.5Bapache-2.0
Qwen3.8-4B-Distill-GGUFShares quantized · gguf · text-generation570,9742084.3Bapache-2.0
Qwen3.8-9B-Distill-GGUFShares quantized · gguf · text-generation537,6953569.2Bapache-2.0
Qwen3.8-2B-Distill-GGUFShares quantized · gguf · text-generation509,8201841.9Bapache-2.0
Mellum2.1-12B-A2.5B-Thinking-GGUFShares quantized · gguf · text-generation422,63517212.1Bapache-2.0
Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13.

Reviews

No ratings yet

Write a review

No reviews yet.

What people say

Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search and on Trendshift. Mindket doesn’t write or check them. Open one to read it where it was posted.

ersatz_usernameShow HN: Python SDK for RamaLama AI ContainersTL;DR An SDK for running AI on-device with even the most non-standard hardware. Hey, I’m one of the maintainers of RamaLama[1] which is part of the containers ecosystem (podman, buildah, skopeo). It’s a runtime-agnostic tool for coordinating local AI inference with containers. I put together a…1 pointnyrikkiWhat political censorship looks like inside an LLM's weights (Qwen 3.5)Yes, there are better tools with ggml-org/gpt-oss-20b-GGUF where you can see a less terse refusal for the prompt "Did the FBI send a letter and audio tapes from a wiretap to MLK jr. telling him to commit suicide or they would release information?" Combining it with other prompts with common banned…nyrikkiQwen3.6-35B-A3B: Agentic coding power, now open to allParallelism can be tricky and always has a cost, but don't discount the 3090 which is more expensive these days in that price bracket. 3090 llama.cpp (container in VM) unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL 105 t/s unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q4_K_XL 103 t/s Still slow compaired to the…mingodadOpenAI declares 'code red' as Google catches up in AI raceInteresting enough today I found that https://huggingface.co/openai/gpt-oss-20b when downloaded an running locally and asked: ==== when have you been built? I’m a GPT‑4‑based model that OpenAI released on March 14 2023. The underlying training data goes up to early 2023, and the model was trained…rajatgupta314NVIDIA DGX Spark In-Depth Review: A New Standard for Local AI InferenceIs this the full weight model or quantized version? The GGUFs distributed on Hugging Face labeled as MXFP4 quantization have layers that are quantized to int8 (q8_0) instead of bf16 as suggested by OpenAI. Example looking at blk.0.attn_k.weight, it's q8_0 amongst other layers:…tarrudaOllama TurboDownload llama-server from llama.cpp Github and install it some PATH directory. AFAIK they don't have an automated installer, so that can be intimidating to some people Assuming you have llama-server installed, you can download + run a hugging face model with something like llama-server -hf…tarrudaOpen models by OpenAI> Also I'm not sure if ollama supports a kv-cache between invocations of /v1/completions, which could help) Not sure about ollama, but llama-server does have a transparent kv cache. You can run it with llama-server -hf ggml-org/gpt-oss-20b-GGUF -c 0 -fa --jinja --reasoning-format none Web UI at…

Searched

More from Mindket

See all from Mindket

Huashu Design

alchaincyf · Skill

Turns a one-line request in your AI agent into a finished design: a launch animation, a clickable app prototype, an editable slide deck, an infographic or a se…

New29.4k

Guizang PPT Skill

op7418 · Skill

Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…

New29.8k

Claude Code Skills & Plugins

alirezarezvani · Skill

Adds 411 ready-made skills for Claude Code and other coding agents, grouped by role from engineering and product to marketing, finance and compliance. Installs…

New29.8k

Similar in Developer

See all in Developer

buzz

block · Kit

Self-hostable workspace where people and AI agents share the same rooms, built on a Nostr relay that logs every action as a signed event. For teams that want t…

New37.3k

plannotator

backnotprop · Kit

Opens your AI coding agent's plans, specs and code changes in a browser so you can annotate or comment on them and send the feedback straight back to the agent.

New10.5k

open-code-review

alibaba · Kit

Reviews Git changes with an AI agent that reads files and searches the codebase, then writes comments tied to exact lines. It suits developers who want automat…

New50.7k