Skip to content
ElastixAI
← All models

OpenAI

gpt-oss-20b

gpt-oss-20b
Open ElastixAI chat ↗

The smaller of OpenAI's two open-weight mixture-of-experts models. 21B parameters with 3.6B active, so it answers at a fraction of the cost of the 120b and fits on a single processor. Reasoning is always on at an effort level you set per request, and tool use is native.

Input
Text
Context window
131K tokens
Precision
4 bits
Price per 1M tokens
$0.025 in · $0.149 out · $0.02 cached
Integration
OpenAI-compatible API
ReasoningConfigurable effortMoE

Playground

gpt-oss-20b

Ready for your prompt

What would you like to explore?

Ask a question, work through an idea, or try a coding task.

Your conversation will appear here.

Limited free previewContinue in chat ↗

Use it in your app

API documentation ↗

Client Libraries

curl https://api.elastix.ai/v1/chat/completions \
  -H "Authorization: Bearer $ELASTIX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-oss-20b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'