Back to models

gpt-oss-120bMos

T powerful open-weight model, fits into an H100 GPU

Context window131,072 tokens
Max output tokens131,072 tokens

Pricing

TypeInputOutput

Parameters

ParameterDefaultRange
Max stop sequences0

Supported features

Capabilities

  • Streaming
  • System Prompt

Endpoints

  • Chat Completions
  • Batch

Use cases

  • text-chat

Model specifications

Model IDgpt-oss-120b
FamilyOther
Access typeOpen Weights
Input modalitiestext
Output modalitiestext
Last updatedJanuary 3, 2026