Instructions to use axteam/gpt-oss-20b-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use axteam/gpt-oss-20b-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf axteam/gpt-oss-20b-GGUF:MXFP4

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf axteam/gpt-oss-20b-GGUF:MXFP4

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4
# Run inference directly in the terminal:
./build/bin/llama cli -hf axteam/gpt-oss-20b-GGUF:MXFP4

Use Docker

docker model run hf.co/axteam/gpt-oss-20b-GGUF:MXFP4

Llama new
LM Studio
Jan

vLLM

How to use axteam/gpt-oss-20b-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "axteam/gpt-oss-20b-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "axteam/gpt-oss-20b-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/axteam/gpt-oss-20b-GGUF:MXFP4

Ollama
How to use axteam/gpt-oss-20b-GGUF with Ollama:
```
ollama run hf.co/axteam/gpt-oss-20b-GGUF:MXFP4
```
Unsloth Desktop

How to use axteam/gpt-oss-20b-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "axteam/gpt-oss-20b-GGUF:MXFP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use axteam/gpt-oss-20b-GGUF with Docker Model Runner:
```
docker model run hf.co/axteam/gpt-oss-20b-GGUF:MXFP4
```

Lemonade

How to use axteam/gpt-oss-20b-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull axteam/gpt-oss-20b-GGUF:MXFP4

Run and chat with the model

lemonade run user.gpt-oss-20b-GGUF-MXFP4

List all available models

lemonade list

Hermes Agent

How to use axteam/gpt-oss-20b-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default axteam/gpt-oss-20b-GGUF:MXFP4

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use axteam/gpt-oss-20b-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf axteam/gpt-oss-20b-GGUF:MXFP4

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "axteam/gpt-oss-20b-GGUF:MXFP4" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

gpt-oss-20b · GGUF (MXFP4)

GGUF build of openai/gpt-oss-20b for llama.cpp / Ollama / LM Studio. Converted by axteam with llama.cpp. The chat template (OpenAI harmony format) is embedded in the GGUF.

File	Quant	Size	Notes
`gpt-oss-20b-GGUF-MXFP4.gguf`	MXFP4	13.8 GB	native 4-bit MoE weights (as released by OpenAI)

gpt-oss-20b is natively MXFP4 — the expert weights ship in 4-bit, so this single file is the faithful, full-quality build (no separate Q4/Q8 needed). OpenAI states gpt-oss-20b runs within 16 GB of memory; it has not been run on a GPU here. Architecture: gpt-oss · Mixture-of-Experts (~21B total, ~3.6B active). 128k context, adjustable reasoning effort.

Quantization method

This is post-training quantization (PTQ): OpenAI released gpt-oss-20b natively in MXFP4, so llama.cpp repacks those 4-bit expert weights into GGUF as-is — no re-quantization and no retraining (so it is not QAT). The precision matches OpenAI's original release.

Requirements


llama.cpp	build ≥ b6325 — architecture `gpt-oss` needs ≥ b6096, and the commands on this card use `-fa on` (≥ b6325). Load-checked here with b12005 (commit e0f142d) and b10724 on CPU — the loader accepts the model, tokenizer and the `-fa on` / q8_0 KV-cache flags; no generation was run in that check.
Memory	the chosen GGUF file (sizes in the table) plus KV cache for the context (`-c`); `--cache-type-k/v q8_0 -fa on` halves the KV cache.
Input	text.
Ollama / LM Studio	not tested here.

Run with llama.cpp

Use a recent llama.cpp build (gpt-oss support, Aug-2025 or later).

# chat server (harmony template is embedded; --jinja uses it)
llama-server -m gpt-oss-20b-GGUF-MXFP4.gguf -c 8192 --jinja
# single answer and exit (-st)
llama-cli   -m gpt-oss-20b-GGUF-MXFP4.gguf --jinja -p "Explain MoE in one sentence." -st

Reasoning effort can be set in the system prompt (Reasoning: low|medium|high).

Reduce KV-cache memory (optional, run-time)

KV-cache type is a run-time choice in llama.cpp (not baked into the file):

llama-server -m gpt-oss-20b-GGUF-MXFP4.gguf -c 8192 --jinja \
  --cache-type-k q8_0 --cache-type-v q8_0 -fa on

Ollama

ollama run hf.co/axteam/gpt-oss-20b-GGUF

License

Apache-2.0, following the original openai/gpt-oss-20b. Weights, tokenizer, and chat template belong to OpenAI. This repo only re-packages them as GGUF.

한국어 요약

openai/gpt-oss-20b의 GGUF 변환 빌드 (llama.cpp·Ollama·LM Studio용) · 제작 axteam

개요

항목	내용
원본 모델	`openai/gpt-oss-20b`
형식	GGUF (llama.cpp)
변환 도구	llama.cpp `convert_hf_to_gguf.py` (OpenAI 원본 MXFP4 가중치를 그대로 옮김, `llama-quantize` 미사용)
모델 구조	MoE (총 약 21B 중 약 3.6B 활성) · 컨텍스트 128k
파일 수	1개

양자화 상세

양자화 절차: PTQ, OpenAI가 MXFP4로 배포한 4비트 전문가 가중치를 재양자화 없이 GGUF로 옮김 (재학습 없음)
chat template: GGUF 파일에 내장 (OpenAI harmony 형식, --jinja로 사용)
품질: 미측정 — 재양자화 없이 OpenAI 배포본과 같은 정밀도

제공 파일

파일	크기	설명
`gpt-oss-20b-GGUF-MXFP4.gguf`	13.8 GB	OpenAI가 배포한 그대로의 4비트 MoE 가중치

요구사항

llama.cpp: b6325 이상 (gpt-oss 구조 지원 b6096부터, 예시 명령의 -fa on 옵션 b6325부터)
로딩 확인: b12005(커밋 e0f142d)와 b10724에서 CPU로 모델·토크나이저 로딩과 -fa on·q8_0 KV 캐시 옵션 인식만 확인 (생성은 실행 미확인)
메모리: 파일 13.8 GB + 컨텍스트 길이(-c)에 비례하는 KV 캐시 — OpenAI 안내 기준 16 GB 메모리에서 실행 가능 (GPU 실행 미확인, KV 캐시를 줄이는 방법은 아래 사용법 참고)
입력: 텍스트
Ollama·LM Studio: 제작 측에서 실행 미확인

사용법

파일 받기: hf download axteam/gpt-oss-20b-GGUF gpt-oss-20b-GGUF-MXFP4.gguf --local-dir . (아래 -m 명령에 필요)
서버 실행: llama-server -m gpt-oss-20b-GGUF-MXFP4.gguf -c 8192 --jinja
한 번 질문 (답변 후 종료): llama-cli -m gpt-oss-20b-GGUF-MXFP4.gguf --jinja -p "Explain MoE in one sentence." -st
추론 강도: 시스템 프롬프트에 Reasoning: low|medium|high 지정
Ollama: ollama run hf.co/axteam/gpt-oss-20b-GGUF
KV 캐시 절감(선택): --cache-type-k q8_0 --cache-type-v q8_0 -fa on — KV 캐시 약 절반, GGUF 파일과 상관없이 실행 시 지정

라이선스

라이선스: Apache-2.0 — 원본 모델의 라이선스를 그대로 따름
권리: 원본 가중치·토크나이저·chat template은 OpenAI 소유, 이 리포는 GGUF로 다시 묶은 것

Downloads last month: 937

GGUF

Model size

21B params

Architecture

gpt-oss

Hardware compatibility

4-bit

Model tree for axteam/gpt-oss-20b-GGUF

Base model

openai/gpt-oss-20b

Quantized

(258)

this model

Collection including axteam/gpt-oss-20b-GGUF

Gpt-OSS

Collection

3 items • Updated 5 days ago