How much VRAM does gpt-oss-120b need?

About 82.5 GB at Q4_K_M, 143.8 GB at Q8_0 and 269.7 GB at FP16, estimated for a ~4k context.

How fast is gpt-oss-120b on an RTX 4090?

We do not estimate tokens/sec when the model does not fit. CPU offload is possible but slow.

What context length can I use?

The weights already do not fit, so longer context is not the limiting factor.

Home Can I run it?gpt-oss-120bgpt-oss-120b on an RTX 4090

No. gpt-oss-120b at Q4_K_M needs about 82.5 GB, and the GeForce RTX 4090 has 24.0 GB usable at a ~4k context.

Can I run gpt-oss-120b on an RTX 4090?

gpt-oss-120b at Q4_K_M needs about 82.5 GB, and the GeForce RTX 4090 has 24.0 GB usable. That figure is 24 GB of VRAM on the spec sheet, of which 24.0 GB is usable for weights and cache, at 1008 GB/s memory bandwidth. OpenAI’s open 120B-class gpt-oss model. It needs large unified memory at Q4, with a much smaller active set per token. These are estimates from our formula, not measured benchmarks.

Quant that fits: None of Q4, Q8 or FP16 needs 82.5 GB
Memory needed: 82.5 GB at Q4_K_M, ~4k context
Usable VRAM: 24.0 GB of 24 GB VRAM
Est. tokens/sec: n/a does not fit

gpt-oss-120b · GeForce RTX 4090 · Everything this machine runs · Q4 vs Q8 vs FP16

Memory at each quant

Quant	Need vs 24.0 GB	Verdict
Q4_K_M	82.5 GB	Q4Won’t fit
Q8_0	143.8 GB	Q8Won’t fit
FP16	269.7 GB	FP16Won’t fit

Estimated tokens per second

No speed estimate: gpt-oss-120b does not fit in the 24.0 GB this machine can use, so we do not publish a tokens/sec figure for CPU offload.

Context headroom

Even at a short ~4k context, the weights do not fit. Extending context would only increase the gap.

Context headroom for gpt-oss-120b on an RTX 4090 at Q4_K_M
Context	KV cache (FP16)	Total at Q4_K_M	Verdict
4k	0.2 GB	82.5 GB	Won’t fit
8k	0.3 GB	82.6 GB	Won’t fit
16k	0.6 GB	82.9 GB	Won’t fit
32k	1.2 GB	83.5 GB	Won’t fit
64k	2.4 GB	84.7 GB	Won’t fit
128k	4.8 GB	87.1 GB	Won’t fit

From the model’s config. Try other sizes in the VRAM calculator or see the context length guide.

What to do instead

Run R1 Distill 32B instead (23.8 GB at Q4). It is a different model, not a drop-in, but it does fit this machine.
To hold gpt-oss-120b at Q4, you need about 82.5 GB. The MacBook Pro M5 Max 128GB has 96.0 GB usable ($6699 Apple list price, 14-inch M5 Max 18C CPU/40C GPU, 128GB/2TB).
CPU offload can load the weights anyway, but it is slow. We do not publish an offload tokens/sec figure.
See the best GPU for this model family and the gpt-oss-120b model page.

Other hardware for gpt-oss-120b

gpt-oss-120b on an RTX 5090Won’t fit
gpt-oss-120b on an RTX 3090Won’t fit
gpt-oss-120b on dual RTX 3090sWon’t fit
gpt-oss-120b on a Mac mini 64GBWon’t fit
gpt-oss-120b on a Mac Studio 256GBFits
gpt-oss-120b on a Mac Studio 512GBFits

Other models on an RTX 4090

Quick answers

Can I run gpt-oss-120b on an RTX 4090?: No. gpt-oss-120b at Q4_K_M needs about 82.5 GB, and the GeForce RTX 4090 has 24.0 GB usable.
How much VRAM does gpt-oss-120b need?: About 82.5 GB at Q4_K_M, 143.8 GB at Q8_0 and 269.7 GB at FP16, estimated for a ~4k context.
How fast is gpt-oss-120b on an RTX 4090?: We do not estimate tokens/sec when the model does not fit. CPU offload is possible but slow.
What context length can I use?: The weights already do not fit, so longer context is not the limiting factor.