Skip to main content

​
About the Provider

OpenAI is the organization behind GPT OSS 20B. They are a major AI research lab and platform provider known for creating influential generative AI models (like the GPT series). With GPT-OSS, OpenAI is extending its technology into the open-source ecosystem, empowering developers and enterprises to run powerful language models without proprietary restrictions.

​
Model Quickstart

This section helps you quickly get started with the openai/gpt-oss-20b model on the Qubrid AI inferencing platform. To use this model, you need:
  • A valid Qubrid API key
  • Access to the Qubrid inference API
  • Basic knowledge of making API requests in your preferred language
Once authenticated with your API key, you can send inference requests to the openai/gpt-oss-20b model and receive responses based on your input prompts. Below are example placeholders showing how the model can be accessed using different programming environments.
You can choose the one that best fits your workflow. 
import requests
import json
from pprint import pprint

url = "https://platform.qubrid.com/api/v1/qubridai/chat/completions"
headers = {
"Authorization": "Bearer <QUBRID_API_KEY>",
"Content-Type": "application/json"
}

data = {
"model": "openai/gpt-oss-20b",
"messages": [
  {
    "role": "user",
    "content": "Explain quantum computing to a 5 year old."
  }
],
"temperature": 0.7,
"max_tokens": 4096,
"stream": False,
"top_p": 0.8
}
response = requests.post(
  url,
  headers=headers,
  json=data, 
)
content_type = response.headers.get("Content-Type", "")
if "application/json" in content_type:
  pprint(response.json())
else:
  for line in response.iter_lines(decode_unicode=True):
      if not line:
          continue

      if line.startswith("data:"):
          payload = line.replace("data:", "").strip()

          if payload == "[DONE]":
              break

          try:
              chunk = json.loads(payload)
              pprint(chunk)
          except json.JSONDecodeError:
              print("Raw chunk:", payload)

​
Model Overview

GPT OSS 20B is a large language model optimized for low-latency inference, local deployments, and specialized use cases. It provides strong reasoning capabilities with adjustable reasoning depth, making it suitable for applications that require transparency, control, and efficient execution without large GPU infrastructure.

​
Model at a Glance

FeatureDetails
Model IDopenai/gpt-oss-20b
ProviderOpenAI
ArchitectureCompact Mixture-of-Experts (MoE) with SwiGLU activations, Token-choice MoE, Alternating attention mechanism
Model Size20.9B Params
Parameters4
Context Length131.1k Tokens
Training DataComprehensive safety evaluation and testing protocols, Global community feedback integration

​
When to use?

You should consider using GPT OSS 20B if:
  • You need fast, low-latency inference
  • You want control over reasoning depth
  • Your application benefits from transparent reasoning
  • You are building tool-based or agentic workflows
  • You want to fine-tune on consumer-grade hardware This model is not intended as a lightweight chat model, but as a reasoning-focused inference model.

​
Reasoning Control

GPT OSS 20B allows you to control how deeply the model reasons before responding.
LevelWhat it means
LowFast responses for simple conversations
MediumBalanced speed and reasoning depth
HighDeep, multi-step analysis for complex tasks
You can set the reasoning level directly in the system prompt: 
Reasoning: high

​
Inference Parameters

Parameter NameTypeDefaultDescription
StreamingbooleantrueEnable streaming responses for real-time output.
Temperaturenumber0.7Controls randomness. Higher values mean more creative but less predictable output.
Max Tokensnumber4096Maximum number of tokens to generate in the response.
Top Pnumber1Nucleus sampling: considers tokens with top_p probability mass.

​
Key Features

  1. Low-latency reasoning – Optimized for fast inference while providing strong reasoning capabilities with adjustable reasoning depth.
  2. Adjustable reasoning depth – Allows control over how deeply the model analyzes a problem (low, medium, high) for speed or detailed multi-step reasoning.
  3. Transparency and debugging – Provides full chain-of-thought access, making outputs easier to understand and debug.
  4. Agentic and tool capabilities – Supports function calling, web browsing, structured outputs, and tool-based workflows for advanced applications.

​
Chain-of-Thought Access

The model provides full chain-of-thought visibility, enabling :
  • Easier debugging
  • Better understanding of how responses are generated
  • Increased trust in outputs

​
Tool and Agent Capabilities

GPT OSS 20B supports agentic workflows and tool usage including:
  • Function calling with defined schemas
  • Web browsing using built-in browsing tools
  • Agentic operations such as browser-based tasks
  • Structured outputs

​
Summary

GPT OSS 20B is a low-latency reasoning model designed for efficient inference and local deployments.
  • It provides adjustable reasoning depth to balance speed and analysis.
  • The model exposes internal reasoning for transparency and debugging.
  • It supports agentic workflows, tool usage, and structured outputs.
  • GPT-OSS 20B can be fine-tuned and run on consumer-grade hardware.