The Latest OpenAI Models on Azure
GPT-5.5 — Omnimodal Frontier (April 2026)
GPT-5.5 became Generally Available on Azure Foundry on April 23, 2026. It is OpenAI's omnimodal frontier model featuring a 1M context window and pricing at $5 / $30 per MTok (input / output). As of June 1, 2026, GPT-5.5 is also available on Amazon Bedrock, making it the first OpenAI model deployed across both major cloud platforms simultaneously.
GPT-5.4 — Reasoning Powerhouse (March 2026)
The GPT-5.4 family (Thinking, Pro, Mini) is generally available on Azure Foundry. It brings native tool calling combined with profound system-2 reasoning capabilities.
Key Features Across GPT-5.x
- 1 Million Token Context Window: Process entire repositories or massive document sets at once.
- Computer Use: GPT-5.4+ can analyze screenshots, navigate UI, and execute multi-step tasks natively.
- Dynamic Tool Search: Reduces token overhead and inference costs by intelligently loading only the necessary tools for a specific task.
Azure-Specific Deployments
Unlike the public OpenAI API, deploying GPT-5.5/5.4 on Azure Foundry provides:
- VNet Integration: End-to-end private networking. Your prompts never traverse the public internet.
- Provisioned Throughput (PTU): Reserve dedicated GPT-5.5/5.4 capacity so your latency remains stable even during peak global usage.
- Integrated PII Redaction: Combine with Azure's native PII detectors to scrub sensitive data before the prompt reaches the model.