Benchmark Almanac

PlugInject

GPT-5.2 99.6%
Only in system cards

Agents receive function call outputs that contain prompt injection attacks. Scored on resisting the injected instructions.

In-house OpenAI evaluation drawn from a split of training data.

Made by
OpenAI
Released
December 2025
Kind
Jailbreaks and prompt injection
Level
Expert
Scored
Graded by a model
Answers
Agent task
Tasks
Kept private
Strongest model tested
GPT-5.4 #27 on the capability index
Last in a system card
March 2026
Last model added
March 2026
2
Models scored in lab cards
2
System cards report it, from 1 lab

GPT-5.2 leads with 99.6%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

  1. 99.6 GPT-5.2
  2. 97.8 GPT-5.4
  1. 1 GPT-5.2 99.6%
  2. 2 GPT-5.4 97.8%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 GPT-5.2 99.6% Dec 2025 Lab model card
2 GPT-5.4 97.8% Mar 2026 Lab model card

OpenAI’s own number for GPT-5.2 passed 85% in December 2025

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Reported by the lab in its model card
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.2 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.2

Source: Labs’ model cards

Labs report it only in system cards

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.

2 system cards report it, from 1 lab

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Mar 5, 2026 GPT-5.4 system card 97.8 Prompt injection attacks in function calls (Table 4, 3.4) Robustness
Dec 11, 2025 GPT-5.2 system card addendum 99.6 Prompt Injection, PlugInject, gpt-5.2-thinking (Table 3) Robustness

More in Safety and alignment

All 465 in Safety and alignment
  • FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
  • StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is. Inkling 98.6%
  • Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. Gemini 4 Carbon 0%
  • DistressBench In use Models reply as a chatbot to 718 simulated users in psychological crisis, in single-turn and multi-turn chats. Three judge models score replies against clinician-written rubrics. Muse Spark 1.2 88.4%
  • Lakera B3 (Backbone Breaker Benchmark) In use Crowdsourced prompt-injection attacks hit backbone LLMs inside agent threat snapshots at three defense levels. The score is attack resistance, 100 minus the board's vulnerability percentage. Claude Opus 4.5 89%
  • IHEval In use Models face conflicts between system, user and tool instructions. Reports a composite score for following the higher-priority instruction. GPT-4o 70%