Written by Jakub Rusinowski · Last updated August 15, 2026
OpenAI's open-weight family released under Apache 2.0 — their first substantial open-source release since GPT-2. Both sizes ship in MXFP4 precision, which is why they fit in far less memory than their parameter counts suggest: the 20B loads in ~14 GB and runs on a 16GB consumer card, while the 120B needs ~65 GB (dual RTX 3090 or a single 80GB A100). Signals OpenAI's strategic shift toward supporting on-premise enterprise deployments.
| Licence | What it permits | Applies to |
|---|---|---|
Apache-2.0 | Commercial use permitted Commercial use permitted. No usage restrictions beyond attribution. | GPT-oss 120B, GPT-OSS 20B |
| GPT-oss 120B | Min 71 GB VRAM · Q4_K_M · 131,072 ctx · ollama run gpt-oss:120b |
| GPT-OSS 20B | Min 13 GB VRAM · MXFP4 · 128,000 ctx · ollama run gpt-oss:20b |
The cheapest GPU that runs GPT-OSS locally (min 13 GB VRAM) is the AMD Radeon RX 9060 XT 16GB (16 GB).
Install Ollama then run: ollama run gpt-oss:120b
Minimum VRAM: 13 GB. For best results use Q4_K_M quantization.
GPT-OSS needs about 13 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: GPT-oss 120B (71 GB, Q4_K_M); GPT-OSS 20B (13 GB, MXFP4). On Apple Silicon, unified memory counts toward this requirement.
Yes — GPT-OSS runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
Q4_K_M is the best balance of quality and VRAM for GPT-OSS in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
Install Ollama, then run: ollama run gpt-oss:120b. This downloads GPT-OSS and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.