Notification of Deep Research in Azure AI FOUNDRY-AN API and SDK based on the offer of Advanced Agent Research Research Openii.
Unlock web research automation on a business scale
We are excited today that we can announce a public insight Deep research in Azure Ai Foundry—NA API and software development (SDK) -on the offer of Advanced Agent Research Open and Openai research agent, fully integrated with the Azure corporate agency platform.
With deep research, developers can build that they are deeply planning agents, analyzing and synthesizing information from the entire web-automatic comprehensive research, creating transparent, auditable outputs, and smoothing multi-stage work with other tools and Azure Azur agents.
Ai AI AI and Knowledge Agents: Fulfillment of additional automation limit of research
Generative models AI and large languages have created research and analysis quickly than ever, power solutions such as Chatgpt Deep Research and Researcher in Microsoft 365 Copilot for individuals and teams. These tools transform everyday productivity and document workflows for millions of users.
Since the organization is trying to integrate the next step deep research directly into their business applications, automation of multiple steps and management knowledge on a business scale-it is clear that programmable, combined and auditable research automation need to be programmed.
Here comes the Azure AI and Deep Research foundry: they offer flexibility for the insertion, expansion and organization of first-class research as services across the corporate ecosystem-and connect it to your data and system.
Deep research skills in Azure AI Foundry Service Agent
Foundry Service Agent’s deep research is built for developers who want to move beyond the cottage window. Azure AI Foundry by offering deep research as a tool for a composable agent via API and SDK, allows customers to:
- Research on the PLC website Using the best research model in its class -based class -based class, with every traceable and source extinction.
- Programmatically building agents This can be triggered by applications, workflows or other agents-and you are repeatedly usable, production beam.
- Complex Orchestrate Work Procedures: Create deep research agents with logical applications, Azure features and other Foundry Service connectors to automate, announcements, and more.
- Engure Enterprise Governance: With the safety, compliance and observance of Azure AI Foundry, customers gain full control and transparency over how research is operated and used.
Unlike packaged chat assistants, deep founders can develop with your needs – ready for automation, expandability and integration with future data sources when expanding support.
How does it work: Architecture and Flow agents’ flow
Foundry Service Agent’s deep research is architectural flexibility, transparency and composition – so you can automate research that is as robust as your business requires.
In its core, deep research model, o3-ethux research, orchestration Multi -stage research pipeline It is firmly integrated with a bing search and uses the latest Openai models:
- Clarfieing Inten and the range of tasks:
When a user or downstream application presents a research question, the agent uses the GPT series models included GPT-4O and GPT-4.1 to clarify the question, the collection of additional content of the established and exactly the scope of the research task. This ensures that the agent’s output is accepted and performed by the stand and that each search is optimized for your business scenario. - Web grounding with bing search:
Once the task is scheduled, the agent safely provokes grounding using the Bing search tool to collect a curator set of high quality, recent web data. This ensures that the research model works from the foundations of authoritative, current sources-and hallucinations from obsolete or irrelevant content. - Deep Research Task Realization:
The O3-Deep-Rerearch model starts the research task. This includes thinking, analysis and synthesizing information across all discovered resources. Unlike simple succession, this emphasizes a step to SEP, turns around because it encourages new knowledge and composition to understand years that are sensitive to nuances, ambiguity and data formulas. - Transparency, safety and compliance with regulations:
The final output is the structured message that it documents not only the answer, but also the path of the model’s reasoning, the quotation of the source and any clarification necessary during the session. As a result, each answer is fully audible Muska for regulated industries and cases of using high betting. - Program integration and composition::
With the Deep Research exhibition as API, Azure AI Foundry, it employs research from anywhere – your own business applications, internal portals, workout automation tools, or as part of a larger agent ecosystem. For example, you can run a research agent as part of a chain with multiple agents: one agent performs a dead web analysis, the other generates a Azure image, while the third e-mail result to creators of deciding with logical Azure applications. This composition is a real game preparation: research is not long in a manual, one -time task, but a building block for digital transformation and continuous intelligence.
This flexible architecture means that deep research can be smoothly built into a wide rage of business and applications. Organizations across industries already evaluate how this programmable research can make high -most valuable scenarios from market analysis and competitive intelligence, for extensive analytical and regulatory news.
Prices for deep research (model: O3-ethrough research) is as follows:
- Input: $ 10.00 for 1 million chips.
- Entry to Cacta: $ 2.50 for 1 million chips.
- Exit: 40.00 $ for 1 million chips.
Finding tokens are charged input token prices for the model used. You will be separately incurred a fee for grounding the search for Bing and the basic GPT model, which is used to clarify the questions.
Start with deep research
Deep research It is now available in a limited public insight for Azure AI Foundry Service. We’re starting:
We can’t wait to see the innovative solution you build. Stay tuned to customers’ stories, new features and future improvements that will continue to unlock the next generation of AI company agents.
(Tagstranslate) ai