GPU Rental Prices
SearchCtrl K

gpt-oss-120b VRAM requirements

gpt-oss-120b has 117 billion parameters, as its publisher states. At 4 bits, the weights alone need approximately 58.5 GB. A full inference estimate is unavailable because the published configuration is restricted or uses an attention layout this calculator does not model.

OpenAI's model card gives 117B parameters with 5.1B active, and says the model fits into a single 80GB GPU such as an H100. Its mixture-of-experts weights, most of the model, ship in MXFP4, a packed 4-bit format, so the 4-bit row is closest to the published files. Read the model card.

Checked 21 November 2026. Memory estimates use decimal GB. Quantized weights may add scales, metadata and buffers; a compatible implementation and model licence are still required.

Change the memory assumptions

Weights only
58.5GB
Weights 58.5 GB · cache not modelled · headroom not modelled.

Full inference not modelled; no GPU fit is claimed.

Memory at different precisions

PrecisionWeights GBInference GBLoRA scenario GBFull fine-tune GB
16-bit234.0Not modelled257.12,108.0
8-bit117.0Not modelled140.12,108.0
4-bit58.5Not modelled81.62,108.0

How the estimate is calculated

Weights = parameters × bits ÷ 8. For supported standard attention, KV cache = 2 × layers × KV heads × head dimension × context tokens × batch size × 2 bytes. Inference adds the chosen percentage of runtime headroom to weights and cache. Divide bytes by 1,000,000,000 for GB.

The LoRA scenario keeps the chosen base weights and allows 18 bytes per trainable adapter parameter, plus the activation allowance. Full fine-tuning uses 18 bytes per parameter for a mixed-precision AdamW scenario, plus that allowance. Its total does not fall when you select 4-bit inference weights. These are budget estimates, not measured memory peaks; training context, optimizer, checkpointing and implementation can change them substantially.

Single-GPU candidates for this inference scenario

The offers update with the memory inputs above. Availability not reported means stock has not been confirmed. Memory capacity alone does not establish runtime compatibility or speed.

No GPU recommendation is made without a supported full memory estimate. Compare GPU memory and rental rates.

Price a training run of gpt-oss-120b and check it fits the node · Compare an API bill with self-hosting

Other models from this publisher

Compare your offers

Saved prices keep their original dates. Confirm the configuration, terms and availability at the source.

Select up to three offers to compare their rates and terms.

Type a GPU, a provider, a model or a question.