← All models

GPT-OSS (120B, MoE)

reasoningApache 2.0tools

OpenAI's open-weight 117B-parameter Mixture-of-Experts model (~5.1B active per token). Configurable reasoning effort (low/medium/high), full chain-of-thought access, and native function-calling. Frontier-class on-device reasoning for workstations and Mac Studio-class hardware — needs ~80GB of usable memory.

At a glance

Parameters
117B
Quantization
Q4_K_M
Context window
131K tokens
Approx. size
65.0 GB
Engines
llama.cpp, MLX, Ollama
License
Apache-2.0 (open)
Version
v1.0.0 , released 2026-05-17
Family
gpt-oss

Tuning

Defaults a gezel client applies out of the box, from real evaluation runs against this exact quantization.

reasoning.effort
"medium"

Model behaviors

Client-side behavior modules the catalog enables for this model (reasoning-tag handling, fabrication detection, prompt shaping).

Sources

llama.cpp (GGUF)
bartowski/openai_gpt-oss-120b-GGUF · Q4_K_M
MLX
mlx-community/gpt-oss-120b-MXFP4-Q4
Ollama
gpt-oss:120b
Upstream
https://huggingface.co/openai/gpt-oss-120b
Manifest
View on GitHub

Tags