OpenAI’s GPT-5.5 might be typically accessible in Microsoft Foundry, bringing OpenAI’s newest frontier mannequin to Azure and the enterprise groups constructing brokers for actual manufacturing work.
OpenAI’s GPT-5.5 might be typically accessible tomorrow in Microsoft Foundry, bringing OpenAI’s newest frontier mannequin to Azure and the enterprise groups constructing brokers for actual manufacturing work.
GPT-5.5 continues a transparent development within the GPT-5 sequence. GPT-5 introduced unified reasoning and velocity right into a single system. GPT-5.4 introduced stronger multi-step reasoning and early agentic capabilities for enterprise use. GPT-5.5 advances this arc with deeper long-context reasoning, extra dependable agentic execution, improved computer-use accuracy, and larger token effectivity—designed for sustained, high-stakes skilled workflows.
Highly effective fashions alone aren’t sufficient to operationalize agentic AI at scale. Microsoft Foundry gives the platform layer that turns frontier fashions into usable, governable techniques that allow enterprises to use safety coverage and administration on the platform degree. Foundry is a unified, interoperable atmosphere to construct, optimize, and deploy AI purposes and brokers with confidence. Prospects profit from broad mannequin alternative, open and versatile agent frameworks, native integration with enterprise techniques and productiveness instruments, and enterprise-grade safety, compliance, and governance. When new fashions like GPT-5.5 grow to be accessible, Foundry makes it straightforward to judge, productionize, and scale them with out friction.
What’s new in GPT-5.5
GPT-5.5 is constructed for skilled situations the place precision, reliability, and persistence matter. GPT-5.5 Professional, a premium variant, extends reasoning depth and activity complexity for probably the most demanding enterprise workloads.
- Improved agentic coding and computer-use: Executes multi-step engineering duties end-to-end—holding context throughout massive techniques, diagnosing the basis reason for ambiguous failures on the architectural degree, and reasoning by way of what else within the codebase a repair will have an effect on earlier than making a transfer. It anticipates downstream testing and evaluate necessities while not having to be informed, and navigates software program interfaces with improved precision and extra dependable restoration when execution takes an sudden flip.
- Autonomous execution and analysis depth: Goes past code to deal with the total span {of professional} work—producing polished deliverables like paperwork, spreadsheets, and shows. For research-intensive workflows, GPT-5.5 operates as an energetic collaborator throughout your complete arc from query to output: refining drafts throughout a number of passes, stress-testing analytical reasoning, proposing approaches, and synthesizing throughout paperwork, knowledge, and code to drive work ahead relatively than simply answering it.
- Complicated reasoning and long-context evaluation: Handles in depth paperwork, codebases, and multi-session histories with out dropping the thread.
- Token effectivity constructed for scale: GPT-5.5 reaches higher-quality outputs with fewer tokens and fewer retries—reducing value and latency for manufacturing deployments at scale.
GPT-5.5 is especially properly suited to domains the place the price of imprecision is excessive—akin to software program engineering, DevOps, authorized, well being sciences, {and professional} companies. With GPT-5.5 in Microsoft Foundry, clients can pair OpenAI’s newest frontier mannequin with enterprise-grade infrastructure to place agentic AI into manufacturing.
Microsoft Foundry: The working system for GPT-5.5 brokers at scale
Entry to a frontier mannequin is simply the start line. What we see from clients is that the arduous half isn’t constructing an agent: it’s operating hundreds of them in manufacturing, with actual isolation, id, and governance. That’s the place Foundry Agent Service is available in.
A revolution is unfolding out there. Now, a developer can reliably cause by way of a enterprise drawback with a coding agent—human’s work together with a mannequin doing heavy considering, analysis, asking questions — and the output is a manufacturing agent: a declarative workflow appropriate for a selected activity, and linked to your small business techniques.
These declarative brokers will be outlined in YAML or written in a harness like Microsoft Agent Framework, GitHub Copilot SDK, or nearly any library. With hosted brokers in Foundry Agent Service, LangGraph, Claude Agent SDK, and OpenAI Brokers SDKs all work the identical approach. Engineers can run a single command to land brokers in an remoted sandbox with a persistent filesystem, a definite Microsoft Entra id, and scale-to-zero pricing. Enterprise prepared brokers, at scale, powered by GPT-5.5.
Be taught extra about Foundry Agent Service and Microsoft Agent Framework.
Pricing
| Mannequin | Enter ($/M tokens) | Cached Enter ($/M tokens) | Output ($/M tokens) |
| GPT-5.5 | $5.00 | $0.50 | $30.00 |
| GPT-5.5 Professional | $30.00 | $3.00 | $180.00 |
Get began with Microsoft Foundry
Cease experimenting and begin constructing your subsequent manufacturing AI workloads with GPT-5.5 in Microsoft Foundry.