Skip to content
AI Atlas
ModelActiveOpen weightsIdentity probable

gpt-oss-safeguard-20b

OpenAIfamily · gpt-osshuggingface.co/openai/gpt-oss-safeguard-20b

gpt-oss-safeguard-20b is a safety reasoning model from OpenAI built upon gpt-oss-20b. This open-weight, 21B-parameter Mixture-of-Experts (MoE) model offers lower latency for safety tasks like content classification, LLM filtering, and trust...

Open in Graph
data quality73

Updated 33 min ago · first seen 11 Sept 2026

model_01M294WW9KM7CQ555DQNXZEDRK

Overview

Identity

Canonical model
Yesidentity confidence: mediumOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
Official checkpoints
official_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
Artifacts
None recordedSeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
Provider deployments
3
API aliases
openai/gpt-oss-safeguard-20bIdentifiers under which providers and evaluators refer to this model.
Folded evaluation variants
0Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.

Openness

Open weights— weights downloadable under Apache-2.0; commercial use allowed; redistribution allowed; derivatives allowed; 4 dimensions unknown.

Weights downloadable under a permissive or Creative Commons licence allowing commercial use; code or data may be missing.

  • Weights

    Yes

  • Inference code

    —

  • Training code

    —

  • Training data

    —

  • Dataset

    —

  • Commercial use

    Yes

  • Redistribution

    Yes

  • Derivatives

    Yes

Licence: Apache License 2.0 (permissive · SPDX Apache-2.0 · stated as “apache-2.0”)

dimensions marked null are unknown, not false

Key facts

Release date

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Openrouter id

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Architecture

Architecture

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Model type

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Parameters

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Weights dtype

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

File size

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Library name

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Pipeline tag

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 d agomedium

Hugging Face repo

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Capabilities

Modalities

Modalities
text
Input
text
Output
text

Capabilities

  • Tool calling

    Yes

    OpenRouter public model & pricing listing · T2

  • Structured output

    Yes

    OpenRouter public model & pricing listing · T2

  • Reasoning

    Yes

    OpenRouter public model & pricing listing · T2

  • Vision

    Unavailable

  • Audio

    Unavailable

  • Fine-tuning available

    Unavailable

Context window

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Max output

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Providers & Pricing3

All offers in the price terminal →
Provider deployments, cheapest output first
ProviderContextInput / 1MCached inOutput / 1MNative unitsStatusObservedSource
GroqCloudcheapest outputopenai/gpt-oss-safeguard-20b131.1Kout 65.5K—output_tokens_per_second=1000active29 min agosince 11 Sept 2026console.groq.comT1
OpenAI APIopenai/gpt-oss-safeguard-20b131.1Kout 65.5K$0.037—active5 d agosince 11 Sept 2026openrouter.aiT2
OpenRouteropenai/gpt-oss-safeguard-20b131.1Kout 65.5K$0.037—active27 min agosince 12 Sept 2026openrouter.aiT2

USD per 1M tokens as published by each provider; native units (per-request fees, flex/priority tiers) are kept verbatim. Rows are append-only — every price change is kept in the history below. Cost of a workload →

Price history

Step lines per provider; amber markers are recorded changes. Click a marker or a row for the evidence behind that price.

Output price · USD / 1M tokens 3 providers

Output price history of gpt-oss-safeguard-20b$0$0.10$0.20$0.30$0.40Sept 26Sept 26Sept 26Sept 26Sept 26OpenAI API: first observed → $0.30 · 11 Sept 2026GroqCloud: first observed → $0.30 · 11 Sept 2026OpenRouter: first observed → $0.30 · 12 Sept 2026
  • OpenAI API
  • GroqCloud
  • OpenRouter
  • OpenRouterfirst observed $0.3012 Sept 2026
  • GroqCloudfirst observed $0.3011 Sept 2026
  • OpenAI APIfirst observed $0.3011 Sept 2026

Input price · USD / 1M tokens 3 providers

Input price history of gpt-oss-safeguard-20b$0$0.02$0.04$0.06$0.08$0.10Sept 26Sept 26Sept 26Sept 26Sept 26OpenAI API: first observed → $0.075 · 11 Sept 2026GroqCloud: first observed → $0.075 · 11 Sept 2026OpenRouter: first observed → $0.075 · 12 Sept 2026
  • OpenAI API
  • GroqCloud
  • OpenRouter
  • OpenRouterfirst observed $0.07512 Sept 2026
  • GroqCloudfirst observed $0.07511 Sept 2026
  • OpenAI APIfirst observed $0.07511 Sept 2026

Hardware fit37

Estimated

37 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—12.9 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—12.9 GB est.Yes
Apple M2 Ultra4bit—12.9 GB est.Yes
Apple M1 Ultra4bit—12.9 GB est.Yes
Apple M3 Max4bit—12.9 GB est.Yes
Apple M4 Max4bit—12.9 GB est.Yes
Mac Studio (Apple M5 Max)4bit—12.9 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—12.9 GB est.Yes
Apple M2 Max4bit—12.9 GB est.Yes
Apple M1 Max4bit—12.9 GB est.Yes
Apple M4 Pro4bit—12.9 GB est.Yes
Mac mini (Apple M5 Pro)4bit—12.9 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—12.9 GB est.Yes
Apple M3 Pro4bit—12.9 GB est.Yes
Apple M1 Pro4bit—12.9 GB est.Yes
Apple M2 Pro4bit—12.9 GB est.Yes
Apple M44bit—12.9 GB est.Yes
iMac (Apple M4)4bit—12.9 GB est.Yes
Mac mini (Apple M6)4bit—12.9 GB est.Yes
MacBook Air (Apple M5)4bit—12.9 GB est.Yes
MacBook Pro (Apple M5)4bit—12.9 GB est.Yes
Apple M24bit—12.9 GB est.Yes
Apple M34bit—12.9 GB est.Yes
Apple M14bit—12.9 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB12.9 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB12.9 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB12.9 GB est.Yes
NVIDIA A100 80GB4bit80 GB12.9 GB est.Yes
NVIDIA H100 SXM4bit80 GB12.9 GB est.Yes
NVIDIA H100 NVL4bit94 GB12.9 GB est.Yes
NVIDIA DGX Spark4bit128 GB12.9 GB est.Yes
NVIDIA H2004bit141 GB12.9 GB est.Yes
NVIDIA H200 NVL4bit141 GB12.9 GB est.Yes
NVIDIA B2004bit180 GB12.9 GB est.Yes
AMD Instinct MI300X4bit192 GB12.9 GB est.Yes
AMD Instinct MI325X4bit256 GB12.9 GB est.Yes
NVIDIA DGX B2004bit1,440 GB12.9 GB est.Yes
Assumptions (7)
  • Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead (or the observed artifact file size when one is recorded).
  • bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
  • KV cache: 2 × layers × kv_heads × head_dim × 2 bytes × context × batch when the architecture is known; otherwise 0.5 GB per 8 192 tokens (× batch), independent of architecture (GQA/MLA models need less).
  • A model 'fits' when the estimate is at most the device memory minus 2 GB reserved for the OS and framework.
  • Mixture-of-experts models are estimated on total parameters (all experts must be resident); active parameters are ignored.
  • Device memory uses the largest configuration when several are listed (e.g. Apple silicon tiers).
  • Multi-GPU: device memories are summed; interconnect bandwidth, tensor-parallel replication and pipeline bubbles are not modelled.

Lineage

Open in Graph →
Explicit derived_from / fine_tuned_from / distilled_from relations stated by sources; artifacts collapsed by kind.
ANCESTORS 1gpt-oss-20b20.9Bgpt-oss-20b — 20.9Bgpt-oss-safeguard-20b21.5B params · this modelgpt-oss-safeguard-20b — 21.5B params · this model

Versions & Artifacts0

Version history

Context windowfirst observation only

11 Sept 2026current

Licensefirst observation only

11 Sept 2026current

Max outputfirst observation only

11 Sept 2026current

Opennessfirst observation only

11 Sept 2026current

Parametersfirst observation only

11 Sept 2026current

Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.

Artifacts 0

No artifact (checkpoint, quantisation, conversion or packaging) points to this model yet.

Papers1

Timeline3

Full timeline →
  • Listed by providerModelgpt-oss-safeguard-20bOpenAI

    OpenRouter lists gpt-oss-safeguard-20b at $0.075 in / $0.3 out per 1M tokens

    openrouter
    openrouter.ai
  • Release date changedModelgpt-oss-safeguard-20bOpenAI

    gpt-oss-safeguard-20b: release date changed from 2025-09-18 to 2025-10-29

    Release date18 Sept 2025→29 Oct 2025openrouter
    openrouter.ai
  • Release date changedModelgpt-oss-safeguard-20bOpenAI

    gpt-oss-safeguard-20b: release date changed from 2025-10-29 to 2025-09-18

    Release date29 Oct 2025→18 Sept 2025huggingface
    huggingface.co

Change history

Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.
1 claims · 1 propertiesShow all properties

Weights availableweights_available1

Claim history for Weights available
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentAI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2highderived

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

Provenance

Attributed facts

43

Source tiers

T1T22 / 41

Freshest observation

1 h ago

Conflicts

None

Source documents 5

Source documents
SourceDocumentTypeTierLast observedSnapshots
Groq — docs & pricingconsole.groq.com/docs/models model_docsT1· Official29 min ago30
OpenRouter public model & pricing listingopenrouter.ai/api/v1/models catalogueT2· Quality secondary27 min ago76
Hugging Face Hub (public pages, model cards, papers)huggingface.co/openai/gpt-oss-safeguard-20b/raw/main/README.md model_cardT2· Quality secondary35 min ago1
Hugging Face Hub (public pages, model cards, papers)huggingface.co/openai/gpt-oss-safeguard-20b model_pageT2· Quality secondary1 h ago29
Hugging Face Hub (public pages, model cards, papers)huggingface.co/models?author=openai&p=0&sort=downloads listingT2· Quality secondary2 h ago32

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.

Data quality (73/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →