Not established from the available sources.
OPENAI · MODEL RELEASE TRACKER
GPT-5.6 Cyber
GPT-5.6 Cyber is an OpenAI cybersecurity model that served as the predecessor to GPT-6 Cyber. According to reporting, it scored 32.1% on the ExploitGym benchmark, was previously classified as "High" on OpenAI’s Preparedness Framework, and had an out-of-bounds (boundary-violation) rate of about 5% in safety evaluations.CURRENT SNAPSHOT2/5 DIMENSIONS WITH DATA
The dimensions that change the decision.
Not established from the available sources.
Not established from the available sources.
ExploitGym score32.1% (reported by OpenAI as the model's score on the ExploitGym benchmark)
Availability in AWS GovCloud (US‑East) via Amazon Bedrock to approved customersDaybreak Red brings GPT-5.6 Cyber to approved customers in AWS GovCloud (US‑East) through Amazon Bedrock for authorized cybersecurity work.
VERIFIABLE FACTS
Every value stays attached to a source and date.
AVAILABILITY · Availability in AWS GovCloud (US‑East) via Amazon Bedrock to approved customersDEVELOPER CLAIM
Daybreak Red brings GPT-5.6 Cyber to approved customers in AWS GovCloud (US‑East) through Amazon Bedrock for authorized cybersecurity work.
LIMITATIONS · Use restricted to authorized cybersecurity workDEVELOPER CLAIM
Access is provided to approved customers for authorized cybersecurity work (i.e., use is restricted to authorized cybersecurity purposes).
BENCHMARKS · ExploitGym scoreDEVELOPER CLAIM
32.1% (reported by OpenAI as the model's score on the ExploitGym benchmark)
SAFETY · Rate of boundary violations in safety evaluationDEVELOPER CLAIM
About 5% (reported as the approximate out-of-bounds rate for GPT-5.6 Cyber in OpenAI's tests)
CAPABILITIES · OpenAI Preparedness Framework classificationDEVELOPER CLAIM
Classified as "High" under OpenAI's Preparedness Framework (the predecessor classification before GPT-6 Cyber)
WHAT CHANGED
Stored passport versions, without reconstructed history.
BENCHMARKS · ExploitGym score+ 32.1% (reported by OpenAI as the model's score on the ExploitGym benchmark)
SAFETY · Rate of boundary violations in safety evaluation+ About 5% (reported as the approximate out-of-bounds rate for GPT-5.6 Cyber in OpenAI's tests)
CAPABILITIES · OpenAI Preparedness Framework classification+ Classified as "High" under OpenAI's Preparedness Framework (the predecessor classification before GPT-6 Cyber)