OpenAI logo

Launch

Open reasoning model for agents
By OpenAIgpt-oss:20b20B parameters32K tokens of contextApache 2.0

OpenAI's open MoE for reasoning, tools and structured outputs. The MXFP4 build fits a 24GB consumer GPU.

GPU configurations for GPT-OSS 20B

1x RTX3090

Best value
~27 tok/s
$0.44per hour

1x RTX3090TI

~29 tok/s
$0.53per hour

1x RTX4090

~40 tok/s
$1.33per hour

1x RTX5090

~58 tok/s
$1.78per hour

1x A100

~48 tok/s
$2.50per hour

1x H100

~78 tok/s
$5.92per hour

1x H200

~88 tok/s
$8.26per hour
Pick a GPU
VRAM
RAM
Storage
CPU

Other models in the private loop

FAQ

Frequently asked about renting GPT-OSS 20B

Which GPU does GPT-OSS 20B need?

The entry configuration is 1x RTX3090, which delivers around 27 tokens per second. For more parallel requests and headroom on speed, go with 1x H200.

How much does GPT-OSS 20B cost?

An hour of GPT-OSS 20B runs from $0.44 to $8.26 depending on the GPU configuration. Billing is hourly: you pay for server time, not for tokens.

How do I call GPT-OSS 20B from my code?

Once deployed you get an OpenAI-compatible endpoint: point any OpenAI client at the base_url you receive and pass gpt-oss:20b as the model name. Code written against the OpenAI API needs no changes.

Does the data stay private?

Yes. GPT-OSS 20B runs on a dedicated GPU server inside a private loop: requests and responses never reach the model vendor and are not used for training.

What context window and license does GPT-OSS 20B have?

GPT-OSS 20B holds 32K tokens of context per request and ships under the Apache 2.0 license, so you can use it commercially on your own hosting terms.