GPT OSS 20B (Databricks, EU)

gpt-oss-20b is one of two open-weight reasoning models released by OpenAI on August 5, 2025, sitting alongside the larger gpt-oss-120b in the gpt-oss family. Both models are distributed under the flexible Apache 2.0 license, which means the weights can be freely used, fine-tuned, and deployed, and the 20b variant is specifically positioned for efficient inference on consumer hardware, including edge devices with roughly 16 GB of memory. This makes it well suited to local deployment, rapid prototyping, and scenarios where sending data to a remote API is not desirable.

The model was trained using a combination of reinforcement learning and techniques borrowed from OpenAI's more advanced internal systems, including o3 and other frontier models, giving it a strong foundation in step-by-step reasoning. According to OpenAI, gpt-oss-20b delivers results comparable to OpenAI o3-mini on common reasoning benchmarks while outperforming similarly sized open models, and it shows particular strength on agentic evaluations such as Tau-bench and HealthBench, even surpassing some proprietary models on tool use and few-shot function calling. In practice, it fits workflows that need reasoning quality and reliable tool integration in a lightweight, open package.