gpt-oss is OpenAI's return to open weights — its first since GPT-2. The August 2025 release shipped two MoE reasoning models under Apache 2.0: gpt-oss-120b (117B total, 5.1B active, single-80GB-GPU serving in native MXFP4) and gpt-oss-20b (21B total, runs on a 16GB consumer GPU), both with configurable reasoning effort and 128K context.
Strategically it validated the open-weight thesis from the least likely vendor, and practically the 20B became one of the most-run local models in the world — the default 'serious model on a laptop.' For pure capability the 2026 Chinese frontier models outclass it, but for accessible, US-origin, Apache-2.0 deployment it remains a staple.
gpt-oss specifications
| Vendor | OpenAI (United States) |
|---|---|
| Architecture | Mixture-of-Experts with configurable reasoning effort |
| Total parameters | 117B (gpt-oss-120b) / 21B (gpt-oss-20b) |
| Active parameters | 5.1B / 3.6B |
| Context window | 128K tokens |
| Modalities | text |
| License | Apache 2.0 — permissive, patent-granting, free for commercial use |
| Release date | 2025-08-05 |
| Weights | Hugging Face · GitHub |
Running gpt-oss locally
The 120B runs on a single 80GB GPU (native MXFP4); the 20B runs on a 16GB consumer GPU or a MacBook. Deliberately engineered for accessible deployment.
What gpt-oss is best for
- Local deployment on consumer hardware (20B on a 16GB GPU)
- US-origin open weights under Apache 2.0 for policy-constrained buyers
- Configurable reasoning-effort workloads
Frequently asked questions
Which gpt-oss should I run locally?
gpt-oss-20b for laptops and consumer GPUs (16GB VRAM or Apple Silicon); gpt-oss-120b if you have a single 80GB GPU (A100/H100) — it ships in MXFP4 specifically to fit one card.
Is gpt-oss as good as ChatGPT?
No — OpenAI positioned it below its proprietary flagships, and 2026 open-weight releases (DeepSeek V4, Kimi K2.5, GLM-5.2) also outscore it. Its value is deployment freedom: Apache 2.0 weights from a US vendor that run on hardware you own.