OpenAI

gpt-oss-20b

Generally availableOpen weights131K contextReasoning

Summary

gpt-oss-20b is an open-weight language model from OpenAI, released on 5 Aug 2025. It accepts text and generates text, with a 131K-token context window and up to 131K output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.02 per million input tokens and $0.09 per million output tokens (OpenRouter), across 10 providers we track.

Open GPT reasoning model for self-hosted agents and controllable deployments

Last verified 18 Nov 2026. Every figure on this page links to its source.

Specifications

DeveloperOpenAI
Release date5 Aug 2025 [models.dev]
StatusGenerally available [models.dev]
WeightsOpen weights [Hugging Face]
Licenseapache-2.0 [Hugging Face]
Parameters20.9B [Hugging Face]
Context window131K (131,072 tokens) [models.dev]
Max output131K tokens [models.dev]
Knowledge cutoffJun 2024 [models.dev]
InputText [models.dev]
OutputText [models.dev]
Reasoning modeYes [models.dev]
Tool callingYes [models.dev]
Structured outputYes [models.dev]
Hugging Faceopenai/gpt-oss-20b [Hugging Face]

gpt-oss-20b API pricing by provider

USD per million tokens. Sorted by input price.

ProviderInputOutputCached inputContextSource
NVIDIA NIM
openai/gpt-oss-20b
FreeFree–131Kmodels.dev, 18 Nov 2026
OpenRouter
openai/gpt-oss-20b
$0.02$0.09$0.009131KOpenRouter, 18 Nov 2026
DeepInfra
openai/gpt-oss-20b
$0.03$0.14$0.01131Kmodels.dev, 18 Nov 2026
Vercel AI Gateway
openai/gpt-oss-20b
$0.03$0.14–131Kmodels.dev, 18 Nov 2026
SiliconFlow
openai/gpt-oss-20b
$0.04$0.18–131Kmodels.dev, 18 Nov 2026
Amazon Bedrock
openai.gpt-oss-20b
$0.07$0.30–131Kmodels.dev, 18 Nov 2026
Amazon Bedrock
openai.gpt-oss-20b-1:0
$0.07$0.30–131Kmodels.dev, 18 Nov 2026
Databricks
databricks-gpt-oss-20b
$0.07$0.30–131Kmodels.dev, 18 Nov 2026
Groq
openai/gpt-oss-20b
$0.07$0.30$0.04131Kmodels.dev, 18 Nov 2026
Amazon Bedrock
us-gov.openai.gpt-oss-20b-1:0
$0.08$0.36–128Kmodels.dev, 18 Nov 2026
Hugging Face Inference Providers
openai/gpt-oss-20b
$0.10$0.50–131Kmodels.dev, 18 Nov 2026
Cloudflare Workers AI
@cf/openai/gpt-oss-20b
$0.20$0.30–128Kmodels.dev, 18 Nov 2026

Running gpt-oss-20b locally

Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.

QuantizationWeightsTotal, 8K contextTotal, 32K context
Q4_K_M (4-bit)13 GB14 GB14 GB
Q8_0 (8-bit)22 GB23 GB24 GB
FP16 / BF16 (16-bit)42 GB43 GB43 GB

At 4-bit with an 8K conversation, gpt-oss-20b needs about 14 GB, so it fits on a graphics card with 16 GB or more, such as the NVIDIA GeForce RTX 4060 Ti 16 GB. Check your own hardware, or see Best GPU for local LLMs on our sister site AIPCs.

Frequently asked questions

What is the context window of gpt-oss-20b?

gpt-oss-20b has a context window of 131,072 tokens (131K) and can generate up to 131,072 tokens in a single response, according to models.dev.

How much does the gpt-oss-20b API cost?

gpt-oss-20b is listed from $0.02 per million input tokens and $0.09 per million output tokens (OpenRouter). Prices last verified 18 Nov 2026.

Is gpt-oss-20b open source?

gpt-oss-20b is an open-weight model: its weights are published on Hugging Face as openai/gpt-oss-20b under the apache-2.0 license. Check the license terms before commercial use.

When was gpt-oss-20b released?

OpenAI released gpt-oss-20b on 5 Aug 2025, according to models.dev.

Which providers offer gpt-oss-20b?

We track gpt-oss-20b on 10 providers: NVIDIA NIM, OpenRouter, DeepInfra, Vercel AI Gateway, SiliconFlow, Amazon Bedrock, Databricks, Groq, Hugging Face Inference Providers, Cloudflare Workers AI.

How much VRAM does gpt-oss-20b need?

About 14 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 23 GB at 8-bit. The 4-bit model fits on a graphics card with 16 GB or more, such as the NVIDIA GeForce RTX 4060 Ti 16 GB. These are estimates from the model's published parameter count and architecture.

Similar models from other labs

Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.

ModelLabReleasedContextInput $/1MOutput $/1MWeights
Baichuan M2 32BBaichuan13 Aug 2025131K$0.07$0.07Open
Granite 4.0 H MicroIBM2 Oct 2025131K$0.02$0.11Open
DeepSeek R1 0528 Qwen3 8BDeepSeek29 May 2025128K$0.06$0.09Open
Nvidia Nemotron Nano 9B V2NVIDIA18 Aug 2025131K$0.06$0.23Open
UI-TARS 7BBytedance22 Jul 2025128K$0.10$0.20Open
Gemma 3 12B ITGoogle DeepMind12 Mar 2025131K$0.05$0.10Open

Change history

  • OpenRouter price changed: input $0.03 → $0.02, output $0.13 → $0.09 per 1M tokens
  • Name changed from GPT OSS 20B to gpt-oss-20b
  • DeepInfra price changed: input $0.03 → $0.03, output $0.14 → $0.14 per 1M tokens
  • Reasoning mode changed from no to yes
  • Max output changed from 8K to 131K
  • Context window changed from 131K to 131K
  • Tool calling changed from no to yes
  • Name changed from openai/gpt-oss-20b to GPT OSS 20B
  • Name changed from GPT OSS 20B to openai/gpt-oss-20b
  • Reasoning mode changed from yes to no
  • Max output changed from 16K to 8K
  • Tool calling changed from yes to no
  • Context window changed from 128K to 131K
  • Context window changed from 131K to 128K
  • Max output changed from 131K to 16K
  • Max output changed from 33K to 131K
  • Max output changed from 16K to 33K
  • Max output changed from 33K to 16K
  • Max output changed from 8K to 33K
  • Max output changed from 33K to 8K
  • Max output changed from 16K to 33K
  • Tool calling changed from no to yes
  • Reasoning mode changed from no to yes
  • Name changed from openai/gpt-oss-20b to GPT OSS 20B
  • Context window changed from 131K to 131K

Sources