OpenAI FRONTIER API · Context: Not publicly disclosed in the launch post

OpenAI GPT-6 Cyber

OpenAI cybersecurity model rated Critical under the Preparedness Framework: 48.5% on ExploitGym (vs 32.1% for GPT-5.6 Cyber), 99.93% out-of-scope avoidance without deployment prompts, used to evidence 81 novel high-severity Firefox bugs and 948 Linux privilege-escalation exploits.

Provider
OpenAI (Proprietary — Daybreak Red (approved customers); API and Codex)
Architecture
Cybersecurity specialist frontier model (parameter count undisclosed); Preparedness Framework Critical for cyber
Parameters
Undisclosed
Context Window
Not publicly disclosed in the launch post (Max Output: 0 tokens)
API Pricing
Input: $0.00 / 1M · Cached: $0.00 / 1M · Output: $0.00 / 1M
Hardware Target
API only

Interactive Token Cost Estimator (GPT-6 Cyber)

Est. API Cost (0% Cache): --/mo | With 70% Prompt Cache: --/mo
Compare all 12 models →

Architectural Strengths & Target Workloads

  • 51% relative ExploitGym gain over GPT-5.6 Cyber
  • Step-change scope alignment (99.93% / 100% with deployment prompts)
  • Proven on Firefox and Linux kernel vulnerability campaigns

Primary Engineering Workloads: Authorized Vulnerability Research · Exploit Validation & Patch Drafting · Security Code Review Agents.

Frequently Asked Questions (GPT-6 Cyber)

How much does OpenAI GPT-6 Cyber cost per 1M tokens?

OpenAI GPT-6 Cyber is priced at $0.00 per 1M uncached input tokens, $0.00 per 1M cached input tokens, and $0.00 per 1M output tokens (Daybreak Red API / Codex for approved customers; public per-token pricing not listed in the Nov 5 launch post).

What is the context window of OpenAI GPT-6 Cyber?

It supports up to Not publicly disclosed in the launch post (0 tokens) and a maximum generation output of 0 tokens.

Full TechPulse Engineering Analysis

OpenAI Ships GPT-6 Cyber: ExploitGym Jumps to 48.5%, Scope Alignment Hits 99.93%

A Critical-rated cybersecurity model under the Preparedness Framework finds 81 novel high-severity Firefox bugs and develops exploits for 948 Linux privilege escalations—while staying inside researcher-set bounds.

Read Full Technical Dispatch (6 min read) →
Soft editorial illustration of a stacked stone watchtower with an ochre window and a thin terracotta spire on a grainy parchment field