OpenAI GPT OSS 120B is an LLM built with a Mixture-of-Experts (MoE) architecture and designed for efficient, large-scale reasoning, coding, and agentic tasks. It excels in processing extremely long contexts—up to 128,000 tokens—while maintaining resource efficiency by activating only a small subset of its total parameters per token, making it practical for both research and production environments.

Some other noteworthy features of OpenAI GPT OSS 120B include its suitability for local and private deployments, and its strong performance on tool-use and code generation tasks.

Metric	Value
Parameter Count	117 billion
Mixture of Experts	Yes
Active Parameter Count	5.1 billion
Context Length	128,000 tokens
Multilingual	Yes
Quantized*	Yes
Precision*	MXFP4

*Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers.