← CanIRunLocal · All hardware · All models
No. gpt-oss 20B (MoE, MXFP4) does not fit on M2 8GB: MXFP4 needs 15.1 GB and 5.4 GB is usable.
Hardware: M2 8GB · Model: gpt-oss 20B (MoE, MXFP4)
M2 8GB has 8 GB of unified memory and 100 GB/s of memory bandwidth. M3 8GB has the same memory and bandwidth, so the fit numbers match; this page is M2 8GB. The checker leaves macOS a share of unified memory, so Ollama can use 5.4 GB. gpt-oss 20B (MoE, MXFP4) (20.9B, 3.6B active, OpenAI) at MXFP4 is a 14 GB download and needs 15.1 GB with headroom for about 8k tokens.
| Best quant | MXFP4 |
|---|---|
| Download size | 14 GB |
| Memory needed | 15.1 GB |
| Usable fast memory | 5.4 GB |
| Fit | does not fit |
| Estimated speed | no estimate, it does not fit |
| Command | ollama run gpt-oss:20b |
Smallest hardware that runs it: RTX 3060 8GB at MXFP4 (partial offload (slower), 15.1 GB needed, 7.5 GB usable, about 13 tok/s (estimate)). Can a RTX 3060 8GB run gpt-oss 20B (MoE, MXFP4)?
No smaller OpenAI model in this list fits M2 8GB.
Open M2 8GB 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