← CanIRunLocal · All hardware · All models
No. gpt-oss 120B (MoE, MXFP4) does not fit on RTX 5090 32GB: MXFP4 needs 67.3 GB and 31.5 GB is usable.
Hardware: RTX 5090 32GB · Model: gpt-oss 120B (MoE, MXFP4)
RTX 5090 32GB has 32 GB of VRAM and 1792 GB/s of memory bandwidth. The checker holds back 0.5 GB of VRAM for the driver and assumes 16 GB of system RAM, so fast memory usable by Ollama is 31.5 GB. Partial offload may also use 60% of that system RAM. gpt-oss 120B (MoE, MXFP4) (117B, 5.1B active, OpenAI) at MXFP4 is a 65 GB download and needs 67.3 GB with headroom for about 8k tokens.
| Best quant | MXFP4 |
|---|---|
| Download size | 65 GB |
| Memory needed | 67.3 GB |
| Usable fast memory | 31.5 GB |
| Fit | does not fit |
| Estimated speed | no estimate, it does not fit |
| Command | ollama run gpt-oss:120b |
Smallest hardware that runs it: M2 Max (38c GPU) 96GB at MXFP4 (fits in fast memory, 67.3 GB needed, 72.0 GB usable, about 56 tok/s (estimate)). Can a M2 Max (38c GPU) 96GB run gpt-oss 120B (MoE, MXFP4)?
A smaller OpenAI model that fits this hardware: gpt-oss 20B (MoE, MXFP4) at MXFP4 (fits in fast memory, 15.1 GB). Can a RTX 5090 32GB run gpt-oss 20B (MoE, MXFP4)?
Open RTX 5090 32GB in the checkerEstimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-23.
A project by OnDevice Labs · llms.txt