GPT-5.6 Cyber

Overview

GPT-5.6 Cyber is a cybersecurity model from OpenAI for authorized vulnerability research and security testing. It accepts text and image input, returns text, and supports reasoning, tool calling and structured output. OpenAI describes it for use in explicitly authorized engagements.

It has a 400K-token context window and can generate up to 128,000 output tokens. It accepts text and image input and returns text. Its knowledge cutoff is February 16, 2026, and it was released on August 7, 2026.

Capabilities

GPT-5.6 Cyber supports reasoning, tool calling, structured output and image input, and returns text output. Its listed tools also include web search, file search, a code interpreter, computer use, image generation, patch-based code editing and a hosted shell.

OpenAI documents GPT-5.6 Cyber for authorized vulnerability research and security testing, and lists vulnerability reproduction, penetration testing, red teaming and complex system analysis as uses.

Pricing

GPT-5.6 Cyber costs from $12.50 per 1M input tokens and from $75.00 per 1M output tokens. Cached input reads cost from $1.25 per 1M tokens. Pricing is tiered, and the prices shown are the starting tier. GPT-5.6 Cyber is available through RealRelay's OpenAI-compatible API.

Use cases

Use GPT-5.6 Cyber in authorized engagements for vulnerability research and reproduction, reading code and system descriptions within a 400K-token context window and calling tools to check findings.

OpenAI also documents penetration testing, red teaming and complex system analysis as uses, where structured output helps turn findings into reports. Image input allows screenshots of consoles or network diagrams to be included in the analysis.