GPT-5.6 Cyber
Advanced cybersecurity model for authorized vulnerability research and security testing (approved access only).
Specs
- Context
- 400K
- Max output
- 128K
- Input price
- $12.5/ 1M tokens
- Output price
- $75/ 1M tokens
- Released
- 2026-08-07
Context
- Context window
- 400K
- Max input
- 272K
- Max output
- 128K
Pricing · USD / 1M tokens
- Input
- $12.5
- Output
- $75
- Cache read
- $1.25
- Cache write
- —
Modalities
- Input
- TextImage
- Output
- Text
API
- API types
Reasoning
- Reasoning
- Yes
- Levels
- —
Info
- Status
- Active
- Released
- 2026-08-07
- Knowledge cutoff
- 2026-02-16
Features
- Confirmed
How to call
1 provider
openaimodel = gpt-5.6-cyberreasoning = reasoning.effort
responses
POSThttps://api.openai.com/v1/responses
JSON
Standard format
{
"id": "gpt-5.6-cyber",
"object": "model",
"created": 1786060800,
"owned_by": "openai",
"name": "GPT-5.6 Cyber",
"api": {
"types": [
"responses"
]
},
"limits": {
"context": 400000,
"input": 272000,
"output": 128000
},
"modalities": {
"input": [
"text",
"image"
],
"output": [
"text"
]
},
"reasoning": {
"supported": true,
"efforts": []
},
"pricing": {
"currency": "USD",
"unit": "1M_tokens",
"input": 12.5,
"output": 75,
"cache_read": 1.25,
"cache_write": null
},
"features": [
"tools",
"structured_output",
"streaming",
"caching"
],
"info": {
"status": "active",
"release_date": "2026-08-07",
"knowledge_cutoff": "2026-02-16",
"description": "Advanced cybersecurity model for authorized vulnerability research and security testing (approved access only).",
"docs": "https://developers.openai.com/api/docs/models/gpt-5.6-cyber",
"verified_at": "2026-10-10"
}
}Official sources
4