NewJoin us for the Forge keynote livestream on November 3rd
DocsLog inGet started
OpenAI logo

OpenAI gpt-oss-20b

Ready
model path:accounts/fireworks/models/gpt-oss-20b
Deploy on demand

Welcome to the gpt-oss series, OpenAI’s open-weight models designed for powerful reasoning, agentic tasks, and versatile developer use cases. gpt-oss-20b is used for lower latency, and local or specialized use-cases.

OpenAI gpt-oss-20b API Features

On-demand Deployment

Docs

On-demand deployments allow you to use OpenAI gpt-oss-20b on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits.

Metadata

State
Ready
Created on
4 AUG, 2025
Kind
Base model
Provider
OpenAI
Huggingface
openai/gpt-oss-20b

Specifications

Calibrated
No
Mixture-of-Experts
Yes
Parameters
20.9B

Supporting Functionality

Fine-tuning
Not Supported
Serverless
Not Supported
Context Length
131k tokens
Function Calling
Not Supported
Embeddings
Not Supported
Rerankers
Not Supported
Supporting image input
Not Supported

gpt-oss-20b FAQs

gpt-oss-20b is an open-weight 21.5B parameter model developed by OpenAI. It is part of the “gpt-oss” series, optimized for lower latency and local or specialized tasks. The model was trained using OpenAI’s Harmony response format and supports configurable reasoning depth for agentic applications.