Recommended Free Tools
Microsoft did not launch 50 standalone AI products at Build 2025. The company announced more than 50 items—including tools, APIs, platform features, protocols, integrations and research projects—centered on its vision of an “open agentic web.”
That vision is a web where AI agents can find information, use software and complete permitted tasks. For businesses, developers and technology buyers, the practical question is not how many tools Microsoft announced, but which parts are usable, which remain in preview, and where each fits.
As an Amazon Associate I earn from qualifying purchases.
Build 2025 began on May 19, 2025. Microsoft’s official overview is available in its Build announcement.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What Microsoft actually announced
Microsoft’s “more than 50 announcements” covered a broad agent-development strategy rather than a catalog of 50 equivalent applications. The announcements addressed five connected layers:
#1 Best Overall
- With 16 GB of memory, runs as many programs as you want without losing the execution
- The 13.5" 2256 x 1504 screen provides a great movie watching experience
- 512 GB SSD is enough to store your essential documents and files, favorite songs, movies and pictures
- 8 Hours battery run time helps you stay unwired and work longer non-stop
- Creation: Copilot Studio, Microsoft 365 Agents Toolkit and SDK, Microsoft Foundry, GitHub, Visual Studio and Visual Studio Code.
- Coordination: Multi-agent orchestration, Agent2Agent support and Microsoft’s agent frameworks.
- Tools and data: Model Context Protocol (MCP), Microsoft 365 Copilot APIs, Azure services, Windows APIs and NLWeb.
- Deployment: Microsoft 365, Teams, websites, Azure, Windows 11 and enterprise applications.
- Governance: Identity, permissions, auditing, isolation and compliance controls.
The announcement therefore described an ecosystem taking shape—not 50 new products that every customer could immediately download.
What “the open agentic web” means
A conventional webpage mainly presents information to people. An agent-ready website could expose structured, semantic and conversational access to that information, allowing an AI agent to discover what the site offers, retrieve relevant content and—in carefully controlled cases—invoke an action.
For example, an agent might search a travel site for flights, query a retailer’s product catalog or retrieve documentation without relying solely on a human-oriented page layout. A later step, such as booking a ticket or placing an order, would require authentication, authorization and safeguards.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsMicrosoft used the term open agentic web to emphasize interoperability rather than a web controlled by one proprietary assistant. But this remains a strategy and ecosystem effort. It depends on websites, software vendors, identity providers and developers adopting compatible interfaces. Build 2025 did not mean that the web had already been transformed.
NLWeb: making websites easier for agents to use
NLWeb is an open Microsoft project intended to help websites add natural-language interfaces using the site’s own data and a model selected by the developer.
Microsoft compared NLWeb’s potential role in the agentic web with HTML’s role in the traditional web. NLWeb endpoints can also function as MCP servers, giving agents a standardized way to interact with a site’s available capabilities.
Potential uses include:
- Conversational product discovery for ecommerce.
- Natural-language search across documentation and publishing archives.
- Travel, hospitality and event search.
- Customer-support and knowledge-base interfaces.
- Content discovery that understands meaning rather than only keywords.
NLWeb does not automatically turn a website into an autonomous agent. The site still needs reliable data, a model, application logic, authentication, rate limits and policies defining which actions are allowed. Conversational retrieval is also different from an authorized transaction. An agent finding an invoice policy is not the same as an agent approving or paying an invoice.
MCP becomes a common connection layer
The Model Context Protocol, or MCP, is a protocol for connecting AI applications and agents to external tools and data. At Build, Microsoft highlighted first-party MCP support across GitHub, Copilot Studio, Dynamics 365, Azure AI Foundry, Semantic Kernel and Windows 11.
Rank #2
- Brilliant OLED Display – Incredible image quality – The 13" PixelSense touchscreen[1], with optional OLED and HDR[2] tech, gives you sharp detail, smooth scrolling, and colors so richly saturated bringing vivid life into every frame - perfect for work, school, streaming, and creative tasks.
- Up to 15.5 hours of battery life[3] - The new Surface Pro is designed for long days, late nights and everything in between.
- Work at the speed of your ideas – Built with the latest Qualcomm Snapdragon X2 Elite (12 Core) processors, Surface Pro delivers fast, AI‑accelerated performance—making it the most powerful Surface laptop for everything from multitasking to demanding workloads.
- The ports you need – Charge on-the-go, transfer data fast, or create the ultimate desktop set up with two USB-C / USB4[4] ports.
- Built-in AI Companion – Work smarter, create freely, and communicate with confidence—Copilot[5] on Windows 11 is always there to help.
Microsoft and GitHub also discussed work on MCP authorization and an MCP server registry concept. The goal is to make it easier to discover and connect to tools while giving organizations more control over which servers are trusted.
Windows’ announcement described MCP access as off by default in its planned architecture, with least-privilege access and auditable sensitive actions as design principles. Windows MCP support was announced as a private developer preview with selected partners, not as a generally available feature at Build 2025. Details are in the Windows Developer Blog.
MCP is not a security solution by itself. Before connecting an agent to a server, an organization should ask:
- Who operates the server?
- Is access read-only or can it create, delete or transfer something?
- What credentials does it receive?
- Are actions logged and revocable?
- Can sensitive operations require human approval?
- Could webpages, documents or tool responses contain prompt-injection instructions?
Microsoft’s agent tools, organized by audience
| Area | Main announcements | Best suited to | Status at Build 2025 |
|---|---|---|---|
| Web | NLWeb and MCP-enabled web experiences | Website and platform developers | Open project and evolving ecosystem |
| Microsoft 365 | Copilot Tuning, multi-agent orchestration, Agents Toolkit, SDK and APIs | Business developers and IT teams | Mixed GA and preview |
| Azure | Azure AI Foundry agent tools, models, evaluation and deployment | Cloud and AI developers | Mixed availability |
| GitHub | Copilot coding agent and Copilot for Azure | Software developers | Mixed GA and preview |
| Windows | Windows AI Foundry, Windows ML, local models, App Actions and MCP | Windows and device developers | Mixed GA and preview |
| Teams | Teams AI Library, MCP and Agent2Agent support | Collaboration-app developers | Core library GA; selected enhancements preview |
| Research | Microsoft Discovery | Researchers and enterprises | Announced platform; current access requires verification |
Copilot Studio: the low-code route
Copilot Studio is aimed at business users, makers and enterprise teams that want to create agents with limited traditional programming. It supports graphical and natural-language construction, Microsoft 365 integration, Power Platform connectors, Azure AI Search, Microsoft Foundry models and external channels.
It is a practical fit for internal workflow automation and business-facing agents. It is less suitable when a team needs a fully custom runtime, advanced evaluation pipelines or deep infrastructure control.
Copilot Studio can involve Microsoft 365 licensing, standalone licensing, credits and pay-as-you-go billing. Microsoft’s pricing page distinguishes internal Microsoft 365 use from broader standalone deployment; Microsoft 365 Copilot licensing should not be assumed to cover public-facing agents.
Microsoft 365 and Teams development
Copilot Tuning and multi-agent orchestration
Microsoft 365 Copilot Tuning was intended to help organizations create more specialized agents using their own data, workflows and processes. Copilot Studio’s multi-agent orchestration allows multiple agents to combine skills on broader tasks.
These capabilities may help a company separate responsibilities—for example, having one agent retrieve policy information while another prepares a workflow response. They also introduce additional failure points: agents can misunderstand one another, select the wrong tool or loop unnecessarily.
Rank #3
- [This is a Copilot+ PC] — The fastest, most intelligent Windows PC ever, with built-in AI tools that help you write, summarize, and multitask — all while keeping your data and privacy secure.
- [The Power of a Laptop, the Flexibility of a Tablet] — Surface Pro 12” is a 2-in-1 device that adapts to you. Use it as a tablet for on-the-go tasks, prop it up with the built-in kickstand, or attach the Surface Pro Keyboard (sold separately) to turn it into a full laptop.
- [Incredibly Fast and Intelligent] — Powered by the latest Snapdragon X Plus processor and an AI engine that delivers up to 45 trillion operations per second — for smooth, responsive, and smarter performance.
- [All Day Battery Life] — Up to 16 hours of battery life[1] means you can work, stream, and create wherever the day takes you — without reaching for a charger.
- [Brilliant 12” Touchscreen Display] — The PixelSense display delivers vibrant color and crisp detail in a sleek design — perfect for work, entertainment, or both.
Agents Toolkit, SDK and Copilot APIs
The Microsoft 365 Agents Toolkit and SDK target developers who need code-level control while deploying across Copilot, Teams, the web and Azure. Microsoft described the toolkit as supporting testing, debugging and deployment.
Microsoft 365 Copilot APIs provide programmatic access to parts of the Copilot stack, including retrieval, chat and meeting-related capabilities. The retrieval API was described as public preview at Build 2025, with other APIs following. Existing Microsoft 365 permissions and compliance behavior remain important: an API does not give an application a license to ignore tenant access controls.
The Teams AI Library was described as generally available, while MCP and Agent2Agent enhancements were among the preview elements. Agent2Agent is intended to help agents coordinate with other agents, but interoperability still depends on authentication, schemas, permissions and vendor-specific behavior.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Azure: the custom application and infrastructure layer
At Build, Azure AI Foundry provided tools for model selection, agent creation, testing, evaluation, deployment and monitoring. Microsoft now markets the broader platform as Microsoft Foundry, so readers comparing current documentation with 2025 coverage may see both names.
Foundry is the more natural starting point when a team needs:
- Multiple model providers and model catalogs.
- Custom orchestration and retrieval-augmented generation.
- Evaluation and monitoring pipelines.
- Azure deployment and enterprise governance.
- MCP connections to external tools.
Build also highlighted Foundry tooling for Visual Studio Code, including agent creation and testing, model exploration, sample code and MCP server connections. GitHub Copilot for Azure was announced as generally available for VS Code. VS Code for the Web – Azure was described as a public-preview browser development environment with Azure, GitHub Copilot, Agent Mode, the Azure Developer CLI and AI Tools extensions preconfigured.
Developers using an Azure application template could use azd up to deploy and azd down to tear down resources, subject to the template and environment configuration. Azure costs can include models, search, storage, compute, tool calls, monitoring and other services; Foundry is not one universal flat-rate subscription. Microsoft’s current Foundry page describes billing by the individual services and features used.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Microsoft also announced Grok 3 and Grok 3 Mini on Azure AI Foundry at the time. Model availability, pricing, regional access and data-handling terms can differ by provider and should be checked separately.
Rank #4
- A PREMIUM PERFORMANCE LAPTOP — Ready for work, school, and creativity. Built for busy days, big projects, and nonstop multitasking. Run video calls, school and work apps, 20+ browser tabs, and AI tools at the same time without slowing down.
- WITH AI BUILT IN — With a dedicated AI chip (Qualcomm Snapdragon X2 Elite), this Copilot+ PC[5] on Windows 11 helps you work smarter and faster. Prompt, create, and automate with ease - ready for even your most demanding tasks.
- A 13.8" TOUCHSCREEN YOU'LL ACTUALLY USE — Sharp colors, real detail, smooth 120Hz scrolling on the PixelSense touchscreen[1] with LCD display[2]. Tap, scroll, or pinch to zoom - whichever feels right for streaming, editing photos, or daily work.
- 20 HOURS OF BATTERY (LEAVE THE CHARGER) — Up to 20 hours of video playback[3] on a single charge. Work from a coffee shop, take it to class/work, or binge an entire season on a long flight — it'll keep up.
- THE PORTS YOU NEED — Two USB-C / USB4[4] ports for fast charging, big file transfers, or hooking up to three 4K monitors when you want a full desktop. Wi-Fi 7 keeps you online and fast wherever you are.
GitHub Copilot coding agent
GitHub Copilot’s coding agent was designed to work asynchronously on repository tasks such as addressing issues and fixing bugs. A typical workflow is to assign a bounded issue, let the agent inspect and modify the repository, review the resulting changes, run tests and decide whether to merge.
This is an AI coding agent—not a replacement for code review, testing, security analysis or human ownership. It is most appropriate where work can be expressed clearly, tests are reliable, branch protections are enabled and generated changes are reviewed.
Windows AI Foundry and local AI
Windows AI Foundry was presented as a unified Windows AI development platform spanning model selection, optimization, fine-tuning and deployment across local and cloud environments.
Free tools Windows power users keep installed
One-click scans. No signup required.
Build’s Windows announcements included:
- Windows ML as an inference foundation.
- Local and open-source model access.
- Copilot+ PC inbox models and AI APIs.
- LoRA customization for Phi Silica, as announced.
- Semantic search and knowledge-retrieval APIs.
- Retrieval-augmented generation scenarios.
- App Actions for agent-invoked application capabilities.
- A native Windows MCP architecture, registry and MCP servers.
Local inference can improve privacy, latency and offline operation, but it is not automatically cheaper or easier. Model size, optimization, memory, drivers, CPU/GPU/NPU support, Windows version and device compatibility all matter. Hardware and model availability should be verified for the target devices.
Availability: what was GA, preview or still a vision?
Build 2025 mixed several categories of announcement:
- Generally available at announcement: GitHub Copilot for Azure for VS Code and the core Teams AI Library.
- Public preview: VS Code for the Web – Azure and the Microsoft 365 Copilot retrieval API.
- Private preview: Windows MCP with selected partners.
- Open or evolving projects: NLWeb, MCP ecosystem work and registry concepts.
- Longer-term platforms and research: Microsoft Discovery and the broader agentic-web vision.
As of 2026, readers should not assume that a Build announcement remains unchanged. Microsoft has renamed or consolidated parts of its portfolio, including marketing Azure AI Foundry as Microsoft Foundry. Availability may vary by country, Azure region, Microsoft 365 license, tenant configuration, Windows edition, hardware and model provider.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Pricing and the commercial decision
There is no single “Microsoft AI agents” price. Costs can span Microsoft 365 per-user licensing, Copilot Studio credits, Azure model and infrastructure consumption, premium connectors, search, storage, monitoring and third-party model charges.
Price checked August 16, 2026: Microsoft’s Copilot Studio pricing page listed Microsoft 365 Copilot at $30 per user per month, paid yearly, and standalone Copilot Studio at 25,000 Copilot Credits for $200 per pack per month, alongside pay-as-you-go and pre-purchase options. Prices, terms and regional availability can change. Pay-as-you-go requires an Azure subscription, according to the pricing page.
Best Value
- Brilliant LCD Display – The 13" PixelSense touchscreen[1], with LCD and enriched HDR[2] tech, unveils crisper whites, darker blacks, and colors so richly saturated bringing vivid life into every frame – perfect for work, school, streaming and creative tasks.
- Up to 15.5hours of battery life[3] - The new Surface Pro is designed for long days, late nights and everything in between.
- Productivity. All Day. Every Day. – Built with the latest Qualcomm Snapdragon X2 Plus (10 Core) processors, Surface Pro delivers fast, responsive performance with built-in AI acceleration—so you can handle everything from everyday tasks to demanding workloads with ease.
- The ports you need – Charge on-the-go, transfer data fast, or create the ultimate desktop set up with two USB-C / USB4[4] ports.
- Built-in AI Companion – Work smarter, create freely, and communicate with confidence—Copilot[5] on Windows 11 is always there to help.
Microsoft Foundry’s current pricing model is consumption-based by service and feature. Buyers should estimate the complete workflow—not only model tokens—including retrieval, tool calls, hosting, logging, security and human-review operations.
Which Microsoft path fits?
| Need | Likely starting point | Important caution |
|---|---|---|
| Internal Microsoft 365 workflow | Copilot Studio or the Agents Toolkit | Check licenses, permissions, connectors and credit consumption. |
| Custom enterprise AI application | Microsoft Foundry | Budget for Azure services, evaluation, monitoring and governance. |
| Repository maintenance | GitHub Copilot coding agent | Require tests, code review, branch protection and security scanning. |
| Local Windows AI | Windows AI Foundry | Confirm device acceleration, model support and update strategy. |
| Agent-ready public website | NLWeb and MCP | Implement identity, authorization, rate limits, logging and action boundaries. |
| Cross-agent interoperability | MCP and Agent2Agent capabilities | Standards do not eliminate trust, schema or permission problems. |
The unresolved risks
Security and prompt injection
Agents can receive malicious or misleading instructions through webpages, documents, email, search results, retrieved business data or MCP servers. A safer design uses least privilege, scoped credentials, isolated execution, read-only defaults, approval for destructive actions, audit trails, server verification, data-loss prevention and easy revocation.
An agent processing invoices should not receive unrestricted payment credentials. An agent modifying cloud infrastructure should propose a change and pass policy checks before execution. An agent booking travel should clearly show the itinerary, price and final action before purchase.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteReliability and accountability
Agents can select the wrong tool, misread business rules, return stale data, claim an action succeeded when it failed or loop across multiple agents. Teams should define explicit success criteria, use structured tool responses, validate results deterministically, limit retries, collect telemetry and provide a human approval path for consequential actions.
Cost and lock-in
Interoperable protocols can reduce integration friction, but they do not make providers interchangeable. Model behavior, pricing, data policies, authentication and platform-specific features still differ. Organizations should document exit options and avoid allowing a preview API or proprietary workflow to become an unexamined dependency.
Bottom line
Build 2025 was Microsoft’s attempt to connect Copilot, Azure, GitHub, Teams, Windows and web standards into a broader agent platform. The most consequential ideas were not 50 new apps, but the infrastructure around them: NLWeb for agent-ready sites, MCP for tool and data connections, developer environments for building agents, and governance controls for deploying them.
The “open agentic web” is therefore best understood as a direction, not a finished replacement for today’s web. Developers can use parts of the stack now, but every project should separate generally available services from previews, verify current product names and pricing, and keep permissions and human accountability at the center.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
