Best GPU for gpt-oss-120b
The smallest memory budget in our catalog that runs gpt-oss-120b cleanly at 8,192 tokens is 78 GiB net, with MXFP4 (61.7 GiB–63.6 GiB total). Cards at that budget: NVIDIA A100 80GB SXM, NVIDIA H100 80GB SXM. The largest file, MXFP4, is comfortable from 96 GiB net: Ryzen AI Max+ 395.
Which card, by net memory
| Net budget | Devices | Best file that runs | Total at 8,192 | Result | Same file at 32,768 |
|---|---|---|---|---|---|
| 78 GiB | NVIDIA A100 80GB SXM, NVIDIA H100 80GB SXM | MXFP4 | 61.7 GiB–63.6 GiB | Fits | Fits |
| 96 GiB | Ryzen AI Max+ 395 | MXFP4 | 61.7 GiB–63.6 GiB | Comfortable | Comfortable |
Rows start at the first budget where the smallest official file is at least tight and stop at the first where the largest file is comfortable. Devices in one row have the same net budget (capacity minus the catalog's default reserve), so they get the same verdict; choose between them on bandwidth, software support and price, which this page does not rank. llama.cpp, one sequence, fp16 KV cache, typical extras. Unified-memory devices use the reserve in their catalog record. "n/a" in the last column means the model's native window is shorter than 32,768 tokens.
Files this ranking uses
| File | Weights | Source repository |
|---|---|---|
| MXFP4 | 60.77 GiB (65,248,815,744 bytes) | openai/gpt-oss-120b |
Byte counts come from the repository file listings recorded on the model page. Memory estimates are formulas, not measurements on these cards. No speed figure is given: none was measured here.
Data sources
- gpt-oss-120b model card
- config.json — revision b5c939de8f754692c1647ca79fbf85e8c1e70f8a
- safetensors index — metadata.total_size=65248815744
- Method and reserves — catalog dataset 2026-10-02.1; hardware capacities from vendor spec pages listed on each device page