Can I run gpt-oss-120b?

gpt-oss-120b needs about 82.5 GB of memory at Q4_K_M, and the cheapest GPU we track that holds it is the NVIDIA RTX Pro 6000 Blackwell 96GB (96 GB, $8565 list price). 7 of the 82 devices we track run it fully at Q4. These are estimates, not measured benchmarks.

116.8B parameters, by openai. License: apache-2.0. Model card (params from HF API, retrieved 2026-10-19) · How much VRAM for 120B models · Best GPU for gpt-oss · Tokens per second by device · Ollama VRAM (gpt-oss:120b) · GGUF sizes Q3–Q8 (llama.cpp) · Q4 vs Q8 vs FP16 · MoE: 5.1B active per token

Q4_K_M needs

82.5 GB 70.8 GB weights

Q8_0 needs

143.8 GB 124.1 GB weights

FP16 needs

269.7 GB 233.7 GB weights

7 of 82 tracked GPUs and Macs run it fully at Q4.

Memory by context length

Longer chats and documents need more memory for the KV cache. Only 18 of its 36 layers keep a cache that grows with context (18 use a 128-token sliding window), so it stays lean at long contexts. Sizes assume an FP16 KV cache; a Q8 cache roughly halves the cache part.

From the model's config.json (18 full-attention layers × 8 KV heads × 64 head size, retrieved 2026-10-25). Check the model card for its maximum supported context. Try other sizes in the VRAM calculator or compare models in our context length guide.

Compare: gpt-oss-120b vs Llama-3.3-70B-Instruct

Can I run gpt-oss-120b on a specific GPU or Mac?

Direct yes-or-no checks for popular hardware, with the quant that fits and an estimated tokens/sec range.

• Can I run gpt-oss-120b on an RTX 4090?Q4Won’t fit
• Can I run gpt-oss-120b on an RTX 5090?Q4Won’t fit
• Can I run gpt-oss-120b on an RTX 3090?Q4Won’t fit
• Can I run gpt-oss-120b on dual RTX 3090s?Q4Won’t fit
• Can I run gpt-oss-120b on a Mac mini 64GB?Q4Won’t fit
• Can I run gpt-oss-120b on a Mac Studio 256GB?Q8Fits
• Can I run gpt-oss-120b on a Mac Studio 512GB?FP16Fits
• Can I run gpt-oss-120b on an RX 7900 XTX?Q4Won’t fit

Which hardware runs it

We may earn a commission when you buy through these links, at no extra cost to you.