openai/gpt-oss-20b

The smaller model in the gpt-oss family is built for lower-latency, local, and specialized workloads. It has 21 billion total parameters with 3.6 billion active at a time, and the series was trained around OpenAI’s Harmony response format, which the model card says is required for correct operation.

This model is intended for reasoning, tool-oriented work, and agentic applications. Its reasoning effort can be set to low, medium, or high to balance output depth and response time, while parameter fine-tuning supports customization. The Apache 2.0 license also makes it suitable for experimentation, commercial deployment, and other projects that benefit from an adaptable, locally runnable foundation.

SiliconFlowopenai/gpt-oss-20bgpt-oss

Open provider page Provider docs

Quick Info

Provider: SiliconFlow
Model key: openai/gpt-oss-20b
Release date: Aug 13, 2025
Last updated: Nov 25, 2025
Input modalities
Output modalities
Capabilities

Cost

Input token cost: $0.04
Output token cost: $0.18

Limits

Output tokens: 8,000 tokens
Context window: 131,000 tokens

Transparent token rates

Compare openai/gpt-oss-20b pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about openai/gpt-oss-20b

SiliconFlow

Coverage

On June 8, 2026, VentureBeat reported on Harness-1, a 20-billion-parameter open-source AI search agent built on top of OpenAI's gpt-oss-20B by a joint team from the University of Illinois at Urbana-Champaign, UC Berkeley, and the open-source vector database Chroma. The article states Harness-1 scores 73% average on rec The report explains that Harness-1 was evaluated across eight complex search benchmarks spanning open-web searches, SEC financial filings, USPTO patent databases, and multi-hop question-answering tasks requiring reasoning across scattered documents. It also highlights that Harness-1 demonstrates the capabilities of Tin

SiliconFlow

Official sourceBenchmark

Compare Hy3-preview and gpt-oss-20b across performance, cost, capabilities, and real-world use cases. See which model fits your needs.

SiliconFlow

CoverageBenchmark

Fireworks AI's technical deep-dive on OpenAI's gpt-oss release (dated 8/5/2025) explicitly covers both gpt-oss-20b and gpt-oss-120b as OpenAI's first open-source LLMs since GPT-2, framing them as strong reasoning models built on a standard mixture-of-experts transformer architecture. The page states performance lands a The blog details agentic features shared by the gpt-oss family, including support for both built-in tools (code interpreter, browser) and user-provided tools, consistent multi-turn trajectories, and selectable low/mid/high reasoning effort levels analogous to o4-mini-high. It explicitly names the 20B variant as competi

SiliconFlow

Official sourceBenchmark

Compare Qwen3-VL-8B-Instruct and gpt-oss-20b across performance, cost, capabilities, and real-world use cases. See which model fits your needs.