Discover OpenAI's gpt-oss-120b Model

A versatile language model for high-reasoning tasks.

Input: text Output: text Context: 131,072 tokens Release: 2025-08-05

OpenAI's gpt-oss-120b, released in August 2025, is a powerful 117B-parameter Mixture-of-Experts language model. Designed for high-reasoning tasks, it optimizes performance with 5.1B parameters activated per forward pass. This text-to-text model is capable of handling a wide range of applications, from structured output generation to native tool use, including function calling and browsing. With a substantial context window of 131,072 tokens, it offers configurable reasoning depth and full chain-of-thought access, making it versatile for general-purpose production use.

Use Cases

Here are a few ways teams apply OpenAI: gpt-oss-120b in practice—from fast drafting to multimodal understanding. Adapt these ideas to your workflow.

Enhance complex reasoning capabilities

Generate structured text outputs

Optimize performance on single GPU setups

Integrate native tool use in applications

Key Features

A quick look at the capabilities that make this model useful in real projects.

117B-parameter Mixture-of-Experts model

Activates 5.1B parameters per forward pass

Optimized for single H100 GPU with MXFP4 quantization

Configurable reasoning depth and chain-of-thought access

Supports native tool use and structured outputs

Specs

Overview

Vendor
openai

Model ID
openai/gpt-oss-120b

Release
2025-08-05

Modalities & context

Input
text

Output
text

Context
131,072 tokens

Parameters & defaults

Supported parameters: frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p

Defaults: temperature 0.2, top_p 0.95

Benchmark tests: OpenAI: gpt-oss-120b

We ran this model against a few representative prompts to show its range. Review the outputs below and be the judge.

Text

Prompt:

Write 150 words on how AI might positively upend work, leisure and creativity

Artificial intelligence is poised to reshape daily life by enhancing work, leisure, and creativity. In the workplace, AI automates repetitive tasks, freeing employees to focus on strategic thinking, problem‑solving, and interpersonal collaboration. Real‑time data analysis and predictive insights empower teams to make faster, more informed decisions, while personalized learning platforms upskill staff on demand. During leisure, AI curates entertainment tailored to individual tastes, recommending books, music, or travel experiences that surprise and delight. Virtual companions and adaptive games provide immersive, responsive experiences that evolve with the user’s mood and skill level. For creators, generative models act as collaborative partners, suggesting novel melodies, visual motifs, or narrative twists, accelerating the ideation process. By handling mundane chores and expanding imaginative horizons, AI encourages humans to pursue higher‑order pursuits, fostering a culture where productivity, enjoyment, and artistic expression reinforce one another. Society will thrive as AI amplifies human potential across every domain today.

Run this prompt on Upend.AI

Ready to try OpenAI: gpt-oss-120b?

Chat with OpenAI: gpt-oss-120b

up.end

/ˌəpˈend/

verb

To “upend” means to completely disrupt, overturn, or drastically change the established order or structure of something. It implies a significant shift or alteration that can potentially have far-reaching consequences. When something is upended, it is turned upside down or transformed in a way that challenges conventional norms or expectations. The term often carries a sense of innovation, transformation, and sometimes even a hint of upheaval, indicating that the changes are not just minor adjustments but rather a fundamental reimagining of the status quo.

Privacy Terms