GPT-OSS

What GPT-OSS is for

GPT-OSS is a reasoning model: it spends output tokens working through a problem before answering, which makes it slower per useful word and better on problems that need the working shown. OpenAI released it in 2025-08 with 2 sizes (20B, 120B) under Apache 2.0, with a 128K context window. Budget for the extra tokens the thinking phase consumes. The catalogue records its strength as OpenAI's open-weight model with strong reasoning.

Sizes and memory

The 20B needs about 14 GB of weights at Q4_K_M and the 120B about 70 GB, a spread of 56 GB across the family. The catalogue's default pick is 20B. Recommended VRAM adds roughly half again on top of the weights for the KV cache and the runtime.

Size	Weights at Q4_K_M	Recommended VRAM	Smallest card in our guide that holds it
20B	14 GB	21 GB	NVIDIA RTX 4060 Ti 16GB
120B	70 GB	105 GB	Apple M3 Ultra

Hardware that runs GPT-OSS

16 of the 19 cards in our hardware guide hold at least one size of GPT-OSS at Q4_K_M. The cheapest that does is the NVIDIA RTX 4060 Ti 16GB at about $450, running the 20B.

The largest size any of them holds is the 120B at 70 GB, on the Apple M3 Ultra (192 GB at 800 GB/s). Bandwidth, not capacity, sets the speed once a size fits, so a card further down the list with faster memory will generate more tokens per second on the same weights than a larger, slower one.

What people use it for

The catalogue lists 4 recorded uses for GPT-OSS: reasoning, agentic, tools, code. It is tagged reasoning, thinking, tools, agentic, openai.

Where GPT-OSS sits in the GPT-OSS family

GPT-OSS is the only GPT-OSS entry in our catalogue, so there is no in-family alternative to weigh it against; the comparison to make is against the models listed at the foot of this page.

The Apache 2.0 licence puts no usage ceiling on it, which matters if the thing you are building has users. That is the practical difference between this and a vendor community licence.

Context window: 128K

128K tokens is roughly 96,000 words at the usual ratio, which is a small repository or a whole book. Two cautions come with a window this size: the KV cache at full length can rival the weights for memory, and models rarely use the far end of their advertised context as well as they use the near end. Treat it as headroom, not as a promise.

How to run GPT-OSS locally

Our catalogue does not record registry tags, so look the current tag up in the Ollama library or on Hugging Face before pasting these. The commands below are the shape of the workflow, not a copy-and-paste recipe.

Option 1: Ollama (simplest)

# find the tag for GPT-OSS at ollama.com/library
ollama run <tag>

Option 2: Mullama (production)

mullama pull <tag>
mullama run <tag>

Option 3: llama.cpp (CLI)

# download a GGUF from Hugging Face, searching for "GPT-OSS GGUF"
./llama-cli -m model.Q4_K_M.gguf -p "Hello, AI!"

Option 4: Python with Mullama or llama-cpp-python

from mullama import Model, Context
model = Model.load("model.Q4_K_M.gguf", n_gpu_layers=99)
ctx = Context(model, n_ctx=4096)
print(ctx.generate("Hello, AI!", 256))

Sources

Model record from src/data/models.json (2026-06-29); card capacities from src/data/gpus.json (2026-06-29). Upstream: Curated from ollama.com/library + community benchmarks + paperswithcode + huggingface.