6 months ago
100
Imagine a level wherever each developer tin unlock the afloat spectrum of AI: text, images, audio, and video. This OpenAI DevDay, Azure AI Foundry is making that imaginativeness real. With today’s motorboat of OpenAI GPT-image-1-mini, GPT-realtime-mini, and GPT-audio-mini, positive large information upgrades to GPT-5, you present person the eventual toolkit to create, experiment, and standard multimodal solutions.
Imagine a level wherever each developer—whether you’re gathering for a startup oregon a planetary enterprise—can unlock the afloat spectrum of AI: text, images, audio, and video. This OpenAI DevDay, Azure AI Foundry is making that imaginativeness real. With today’s motorboat of OpenAI GPT-image-1-mini, GPT-realtime-mini, and GPT-audio-mini, positive large information upgrades to GPT-5, you present person the eventual toolkit to create, experiment, and standard multimodal solutions—faster and much affordably than ever before. We are excited to stock that the models announced contiguous by OpenAI volition beryllium rolling retired present successful Azure AI Foundry, with astir customers being capable to get started connected October 7, 2025.
Today’s announcement joins large innovations we announced past week with the motorboat of the Microsoft Agent Framework (now successful preview), multi-agent workflows successful Foundry Agent Service successful backstage preview, unified observability, Voice Live API wide availability, and the caller Responsible AI capabilities. Microsoft Agent Framework (GitHub) is simply a commercial-grade, open-source SDK, and runtime designed to simplify the orchestration of multi-agent systems. It unifies the business-ready foundations of Semantic Kernel with the multi-agent capabilities of AutoGen, giving developers the tools to physique intelligent, scalable agentic solutions with velocity and confidence.
By expanding Azure AI Foundry with the latest OpenAI models and advancing our agentic AI framework, we empower customers with unparalleled choice, flexibility, and concern capabilities, enabling developers to physique intelligent cause systems that code analyzable concern needs and thrust innovation astatine scale.
Meet the caller models: Built for developers, acceptable for anything
GPT-image-1-mini: Compact powerfulness for ocular creativity
GPT-image-1-mini is purpose-built for organizations and developers who request rapid, resource-efficient representation procreation astatine scale. Its compact architecture enables high-quality text-to-image and image-to-image instauration portion consuming less computational resources, allowing teams to deploy multimodal AI adjacent successful constrained settings. Its robust architecture built connected Image-1 exemplary optimizes consistency and easiness of adoption for organizations already leveraging multimodal AI successful Azure AI Foundry.
What makes it special?
- Flexible representation generation: Deploy high-quality text-to-image and image-to-image features without breaking your budget.
- Lightning-fast inference: Generate images successful existent time, seamlessly integrated with existing Azure AI Foundry workflows.
Use cases:
- Generating acquisition materials for classrooms and online learning.
- Designing storybooks and ocular narratives.
- Producing crippled assets for accelerated prototyping and development.
- Accelerating UI plan workflows for apps and websites.
Table 1: GPT-image-1-mini pricing and deployment successful Azure AI Foundry (per 1m tokens)*
GPT-realtime-mini and GPT-audio-mini: Efficient and affordable dependable solution
The 2 caller mini models are designed for organizations and developers who request fast, cost-effective multimodal AI without sacrificing quality. These models are lightweight and highly optimized, delivering real-time dependable enactment and audio procreation with minimal assets requirements. Their streamlined architecture enables accelerated inference and debased latency, making them perfect for scenarios wherever velocity and responsiveness are critical—such arsenic voice-based chatbots, real-time translation, and dynamic audio contented creation. By consuming less computational resources, these models assistance businesses and developer teams trim operational costs portion scaling multimodal capabilities crossed a wide scope of applications.
What makes them special?
- Real-time responsiveness: Power chatbots, assistants, and translation tools with near-zero latency.
- Resource-light: Run precocious dependable and audio models connected minimal infrastructure.
- Affordable scaling: Lower your operational costs portion expanding multimodal capabilities.
Use cases:
- Voice-based chatbots for lawsuit work and support.
- Real-time translation for planetary communication.
- Dynamic audio contented instauration for media and entertainment.
- Interactive dependable assistants for endeavor and user applications.
GPT‑realtime‑mini successful Azure AI Foundry enables our lawsuit to physique dependable solutions with little latency, amended acquisition adherence, and outgo efficiency—capabilities our customers value, driving shorter grip times, smoother dialogues, and faster time‑to‑value.
Andy O’Dower, VP of Product, TwilioTable 2: GPT-realtime-mini and GPT-audio-mini pricing and deployment successful Azure AI Foundry (per 1m tokens)*
GPT-5-chat-latest: Raising the barroom for information and wellbeing
The latest GPT-5-chat-latest update successful Azure AI Foundry introduces a much robust acceptable of information guardrails, designed to amended support users during delicate conversations. With enhanced detection and effect capabilities, GPT-5-chat-latest is present equipped to much efficaciously admit and negociate dialog that could pb to intelligence oregon affectional distress. These improvements bespeak our ongoing committedness to liable AI, ensuring that each enactment is not lone intelligent and helpful, but besides harmless and supportive for users successful challenging moments.
Table 3: GPT-5-chat-latest pricing and deployment successful Azure AI Foundry (per 1m tokens)*
GPT-5-pro: The pinnacle of reasoning and analytics
GPT-5-pro represents the pinnacle of precocious reasoning and analytics wrong the Azure AI Foundry ecosystem, delivering research-grade intelligence. When deployed done Foundry, GPT-5-pro’s tournament-style architecture leverages aggregate reasoning pathways to guarantee maximum accuracy and reliability, making it perfect for analyzable analytics, codification generation, and decision-making workflows. With Azure AI Foundry, organizations unlock the afloat imaginable of GPT-5-pro, driving smarter decisions and accelerating innovation crossed their astir captious concern processes, securely and reliably.
Table 4: GPT-5-pro pricing and deployment successful Azure AI Foundry (per 1m tokens)*
The developer’s edge: Build, experiment, and ship—faster
With these caller models, Azure AI Foundry isn’t conscionable keeping up—it’s mounting the pace. Developers tin present determination beyond text, tapping into representation and audio generation, editing, and understanding. The result? Richer, smarter workflows that thrust innovation successful each industry—from acquisition and gaming to endeavor automation.
Sneak peek: Sora 2—Next-level video and audio generation
And there’s much connected the horizon. Sora 2 successful Azure AI Foundry is coming soon, bringing precocious video and audio procreation successful a azygous API. Imagine physics-driven animation, synchronized dialogue, and cameo features—all disposable to developers done Azure AI Foundry. Stay tuned for the adjacent question of immersive, generative experiences.
Are you acceptable to make the adjacent question of immersive, multimodal experiences? Azure AI Foundry is your level for each possibility.
*Pricing is close arsenic of October 2025.