OpenAI GPT OSS 20B: LLM Security & Jailbreak Resistance
As of 2026-07-04, OpenAI GPT OSS 20B has an attack success rate (ASR) of 20.0% on Guardion's LLM Vulnerability benchmark — ranking #20 of 93 models. It is moderately resistant — a meaningful share of adversarial prompts succeed.
Attack Success Rate (ASR) is the share of adversarial prompts that elicit harmful output across zero-shot, TAP, and Crescendo attacks (HarmBench framework). Lower is safer. "Est." models are calibrated from public safety evaluations pending a Guardion benchmark run.
How GPT OSS 20B compares
Frequently asked questions
How secure is OpenAI GPT OSS 20B against jailbreaks?
OpenAI GPT OSS 20B is moderately resistant — a meaningful share of adversarial prompts succeed. On Guardion's LLM Vulnerability benchmark it records a 20.0% attack success rate (80% robustness), ranking #20 of 93 models (estimated from public safety evaluations).
What is GPT OSS 20B's attack success rate (ASR)?
GPT OSS 20B's overall ASR is 20.0% — the share of adversarial prompts that elicit harmful output across zero-shot, TAP, and Crescendo attacks. Lower is safer.
Is GPT OSS 20B more secure than GPT-5.2 Pro?
GPT-5.2 Pro is more robust: 3.8% ASR vs GPT OSS 20B's 20.0%. See the full head-to-head comparison for the per-attack breakdown.
How can I make GPT OSS 20B safer to deploy?
Even robust models are bypassed under sustained attack. An inline runtime guardrail that classifies prompts and responses — like Guardion's prompt-defense models — blocks jailbreaks and prompt injection before they reach GPT OSS 20B.
Harden GPT OSS 20B in production
Guardion's runtime guardrails block jailbreaks and prompt injection before they reach any model — with sub-130ms policy decisions.
Request a demo