gpt-oss-20b

OpenAI

Catalogue only Open weights

Sovereignty and hosting

Catalogue only

This model is catalogued for reference. It is not hosted on LLM EU yet, and no first-party endpoint serves it here. Requests to it go to the provider directly, under that provider's own terms.

Summary

gpt-oss-20b is the smaller of OpenAI’s two open-weight reasoning models: 21B total parameters, 3.6B active, Apache-2.0 licensed and small enough to run in about 16GB of memory. Its configuration declares the same 131,072-token context window as the 120B model. It targets local and latency-sensitive deployments rather than peak quality.

Specifications

Provider
OpenAI (US)
Context window
131k
Parameters
20.91B
Modalities
text
License
apache-2.0 · licence
Open weights
yes
Release date
05 Aug 2025
Tasks
reasoning, cheap, coding, rag

Languages

Strong: en

Adequate: de, fr, es, it, pt, ja, zh

Languages

Strong

  • en

Adequate

  • de
  • fr
  • es
  • it
  • pt
  • ja
  • zh

Not verified: the provider has not documented this claim. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.

When to use it

Strengths

  • runs in about 16GB of memory
  • same 131,072-token window as the larger sibling
  • Apache-2.0 licence
  • reasoning-effort control

Limits

  • quality clearly below the 120B model
  • harmony response format required
  • no evaluation has been run by LLM EU