Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
On-device AI in Windows means that a model performs its inference on your PC, using the processor, graphics processor or neural processing unit (NPU), instead of sending that input to a cloud model. It does not mean that every Windows AI feature is permanently offline: some require an initial model download, online search or a cloud provider for a follow-up action.
Windows has several separate local-AI paths. Copilot+ PC features are built into Windows and tied to specific hardware; applications can run downloaded models through Foundry Local; and developers can deploy ONNX models with Windows ML. The exact feature, Windows build, language, driver and model determine what continues to work without a connection.
What “on-device AI” means in Windows
Inference is the stage at which a trained model analyzes input and produces an output. With on-device inference, that calculation happens on the Windows computer rather than on a remote server. The input might be a camera frame, microphone audio, a document, a screenshot or text selected in an application.
Microsoft’s Windows AI overview describes a mixed architecture: Windows AI APIs for built-in experiences, NPU acceleration, cloud APIs and developer frameworks. Therefore, “AI in Windows” is not one program with one network policy. A single feature can analyze something locally and then offer an optional action that contacts an online service.
#1 Best Overall
- Step Up to Next-Level Performance - Redefine your laptop experience with the Acer Aspire 16 AI. Powered by the Snapdragon X X1-26-100, a premium integrated GPU with up to 1.7 TFLOPs and NPU with 45 TOPs for optimized processing across CPU, GPU, and NPU workloads and delivering best-in-class performance and power efficiency.
- New AI Superpowers - Discover the power of Recall (preview), improved Windows search, and Click to Do (preview) on Copilot+ PCs. Effortlessly locate past content, perform natural searches, and interact with text and images – all while ensuring your data remains private and you stay productive*.
- Built on Brilliant AI Foundations - The Acer Aspire 16 AI harnesses the industry-leading Qualcomm AI Engine with an integrated Qualcomm Hexagon NPU, delivering transformative experiences for creativity, video conferencing, security, and productivity assistants. The Qualcomm AI Engine supports Windows Studio Effects and many other AI-accelerated applications and experiences, to make possibilities endless.
- Screens that Speak to Your Senses - Immerse yourself in a world of vibrant visuals. Enjoy stunning clarity, rich 100% sRGB colors, and sharp detail on the 16" 120Hz WUXGA ultra-high-resolution touchscreen display – acting as a panoramic playground for entertainment, artistic expression, and engaging AI experiences that dazzle the eye.
- Streamline Your Settings with AcerSense - Intelligent Acer AI solutions are at your fingertips. Effortlessly get answers, streamline settings, optimize your video presence, and elevate communication. Experience intuitive AI that’s easy to use and seamlessly enhances your productivity.
“Offline” has a narrower meaning. A feature is offline-ready only when its required software, model and configuration are already on the PC and the specific operation does not call a network service. A local model that has not yet been downloaded cannot run offline, and a local text analysis does not make a subsequent cloud-based rewrite or search offline.
Which Windows AI features can work without internet?
| Feature or route | What runs locally | What must be prepared first | Where internet can still matter |
|---|---|---|---|
| Recall | Snapshots are analyzed and stored on the PC. | Initial setup on a supported Copilot+ PC; documented memory, storage and NPU requirements apply. | Microsoft says internet or a cloud connection is not required or used to save and analyze snapshots. Rollout and availability depend on the Windows build and device. |
| Click to Do text actions | Phi Silica performs supported text actions on a Copilot+ PC. | Supported Copilot+ hardware, Windows feature availability and a supported language. | Some choices made after local analysis send selected content to an online provider. |
| Windows Studio Effects | Supported camera and microphone effects can be applied on the device. | Windows 11 version 22H2 or newer, a supported NPU and an enabled manufacturer driver. | The effects themselves are local; video-call or other application services still use their own network connections. |
| Foundry Local | A downloaded, cached compatible model performs inference locally. | The model must be downloaded while connected and remain cached. | The first model download needs internet. Catalog information may refresh at startup, although cached information can be used while disconnected. |
| Windows ML | An application runs an ONNX model through a selected local execution provider. | The app, ONNX model, runtime and suitable CPU, GPU or NPU must be installed. | There is no single Windows-wide offline guarantee; the application determines whether it downloads models or calls services. |
These are different paths, so a PC can support one and not another. Microsoft’s feature documentation is the authority for current device, build and language eligibility.
Recall: local snapshots with strict hardware requirements
Microsoft’s Recall documentation says snapshots are analyzed and stored on the PC. It also states that no internet or cloud connection is required or used to save and analyze those snapshots. That is a statement about Recall’s snapshot processing, not a promise that every application opened from a result works without a network.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesThe Recall application card lists a Copilot+ PC, a 40 TOPS NPU, 16 GB of RAM and 256 GB of storage among the requirements. It also says at least 50 GB of free storage is needed to enable Recall. These are Microsoft’s documented requirements for this feature, not universal requirements for running local AI on Windows.
What to check before enabling Recall
- Confirm that the exact computer is a Copilot+ PC rather than relying on a generic “AI PC” label.
- Verify the installed Windows release and current manufacturer drivers.
- Check that the machine meets the 40 TOPS NPU, 16 GB RAM, 256 GB storage and 50 GB free-space figures listed by Microsoft.
- Complete Recall’s initial configuration. A feature that has not been configured cannot create a usable local snapshot history.
Microsoft can change supported hardware, rollout status and Windows build requirements, so consult the current management documentation for a particular deployment.
Rank #2
- RELIABLE PERFORMANCE FOR EVERYDAY WORK: The Intel N150 processor works with 8GB LPDDR5 memory and 128GB UFS 2.2 storage to support web browsing, email, document editing, online classes, video streaming, and routine multitasking. Integrated Intel Graphics provides dependable visuals for business, education, and everyday home use.
- CLEAR 15.6-INCH FULL HD DISPLAY: The 1920x1080 anti-glare display offers a spacious view for documents, presentations, research, online learning, and entertainment. Its 250-nit brightness, 88% active-area ratio, and TÜV Rheinland Low Blue Light software solution support comfortable viewing during extended work or study sessions.
- LIGHTWEIGHT AND DURABLE DESIGN: Starting at only 3.42 lbs and measuring 0.70 inches thin, this Arctic Grey Lenovo laptop travels easily between home, school, and the office. MIL-STD-810H testing adds everyday durability, while the full-size keyboard includes a dedicated Copilot key for convenient access to AI assistance.
- MODERN CONNECTIVITY AND PRIVACY: Wi-Fi 6 and Bluetooth 5.2 provide reliable connections for networks and accessories. Two USB-A ports, USB-C with Power Delivery and DisplayPort, HDMI 1.4, an SD card reader, and a 3.5mm audio jack support displays and peripherals, while the 720p camera includes a physical privacy shutter.
- READY FOR BUSINESS AND EDUCATION: Windows 11 Home and Microsoft 365 Personal provide familiar tools for documents, communication, coursework, and daily productivity. A 47Wh battery supports mobile workflows, while the included 65W power adapter enables efficient charging. Dolby Audio stereo speakers and dual-array microphones enhance online meetings and classes.
Click to Do: local text analysis, mixed network actions
On Copilot+ PCs, Click to Do uses Microsoft’s Phi Silica model for certain text operations on the device. In English, Microsoft lists Summarize, Create a bulleted list and Rewrite among the available actions. The retrieved documentation describes a more limited action subset for French and Spanish, so the menu is not identical in every language.
The important boundary is between the analysis and the action that follows it. Microsoft’s Click to Do management documentation notes that some selected choices send the selected content to an online provider. A local summary option and an online provider action should therefore be treated as separate cases. If keeping text on the PC is essential, read the description of each action and avoid choices identified as online.
Why language and build checks matter
- The Phi Silica text-action set is feature-specific; seeing Click to Do does not imply that every action is installed.
- English has the documented set of Summarize, bulleted-list creation and Rewrite; French and Spanish have fewer documented actions.
- Windows updates can change the available menu, supported languages and provider behavior.
Windows Studio Effects: local camera and microphone processing
Studio Effects can process supported camera and microphone effects locally, including background blur, eye contact, auto framing, voice focus, portrait light, creative filters and an eye-contact teleprompter effect. Microsoft’s Studio Effects overview requires Windows 11 version 22H2 or newer, a supported NPU and an enabled manufacturer-installed driver.
How to find the controls
- Open Quick Settings from the taskbar and look for Studio Effects while a supported camera or microphone is available.
- If the control is not shown, open Settings > Bluetooth & devices > Cameras and check whether the camera exposes the feature.
- Install the computer maker’s current camera, audio and NPU drivers, then restart if Windows requests it.
- Test one effect at a time in the calling application. The effect can be local even though the call itself is transmitted through that application’s online service.
A missing Studio Effects panel usually indicates unsupported hardware, an unavailable driver, an older Windows version or a camera that does not expose the required capabilities. It does not by itself prove that the PC has no local-AI ability.
Running your own local models with Foundry Local
Foundry Local is Microsoft’s route for running compatible open-source language models on a Windows computer. According to Microsoft’s getting-started guide, the model must be downloaded and cached before disconnected inference is possible.
Rank #3
- It's possible on your Intel AI PC - Equipped with an Intel Core Ultra 7 processor (Series 2), the Aspire 14 Al brings new AI experiences in productivity, creativity and security through a combination of CPU, GPU and NPU. This combo delivers the speed and responsiveness to handle any task with ease -along with all-day battery life of up to 22 hours and smooth multitasking performance. (Battery life was measured under specific test settings pursuant to video playback scenarios)
- New AI Superpowers - Discover the power of Recall (preview), improved Windows search, and Click to Do (preview) on Copilot plus PCs. Effortlessly locate past content, perform natural searches, and interact with text and images – all while ensuring your data remains private and you stay productive. ( Copilot plus PC experiences vary by device and market and may require updates continuing to roll out through 2025; Recall and Click to Do will be coming to European Economic Area later in 2025; timing varies. See aka.ms/copilotpluspcs)
- Indulge Your Eyes - Immerse yourself in a world of vibrant detail with a breathtaking 14" WUXGA 1920 x 1200 ultra high-resolution display. This expansive, panoramic screen is your canvas for entertainment, artistic creativity, and captivating AI experiences that will leave you in awe.
- Smart and Effortless AI - Intelligent AI solutions are at your fingertips with AcerSense. Streamline settings, optimize your video presence, and elevate communication - all with intuitive AI that’s easy to use and enhances productivity seamlessly. Just press the AcerSense key on the backlit keyboard for instant access and experience the magic of AI
- Style and Substance - The Aspire 14 Al boasts a sleek, durable, and lightweight aluminum chassis, with an ultra-modern design and a 180° lie-flat hinge for versatile and convenient use on the go. Ideal for work, study, or creative pursuits wherever you are.
The offline workflow
- Connect to the internet and install or open the Foundry Local tooling required by the model you intend to use.
- Choose a compatible model and allow its files to download completely. The initial download requires internet access.
- Run a test prompt while connected so you can confirm that the model, runtime and hardware are compatible.
- Disconnect the PC and run the same local workload. Inference can continue from the cached model if the application does not request an online provider or fresh data.
- Reconnect periodically if you need catalog metadata or model updates. Microsoft says catalog information may refresh at startup, while cached information can still be used offline.
A Copilot+ PC is not mandatory for every Foundry Local workload. Microsoft notes that some compatible models have CPU variants, so a suitable CPU can run them without a dedicated GPU or NPU. Compatibility and performance still depend on the selected model and the computer.
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 →Windows ML for developers: local ONNX inference
Windows ML is a developer framework for deploying ONNX models. It can select among available execution providers, allowing an application to use a CPU, GPU or NPU when the model and drivers support it. This is not a consumer feature with one universal offline switch.
For an offline Windows ML application, the developer must package or locally install the model and runtime dependencies, select an execution provider that exists on the target machine and avoid code paths that fetch remote data. An app can use Windows ML for local inference and still require internet for account sign-in, document synchronization, web search or a cloud fallback. The app’s own documentation and network design determine the complete behavior.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Do you need a Copilot+ PC for Windows AI?
Not for every kind of local AI. Microsoft’s Copilot+ PC developer guidance defines Copilot+ PCs as a Windows 11 hardware class with a high-performance NPU and identifies Snapdragon, AMD and Intel device families. The guidance says the NPU enables on-device Copilot+ experiences.
Many new built-in experiences target an NPU rated at 40+ TOPS (more than 40 trillion operations per second). This is a hardware capability threshold for eligible features, not a measurement of real-world model speed, battery life or answer quality. A computer marketed as an “AI PC” may not meet the requirement for Recall, Click to Do or every Studio Effect.
Recommended Free Tools
Rank #4
- 2K IPS TOUCHSCREEN DISPLAY - 1920 x 1200 resolution delivers incredible detail, wide-viewing angles, and lifelike color reproduction
- AMD RYZEN AI 5 430 PROCESSOR - Unlock powerful AI-driven experiences with a Copilot+ PC powered by an AMD Ryzen AI processor designed to enhance creativity, simplify and streamline your day, and give you valuable time back to do more
- ENJOY UP TO 19 HOURS AND 30 MINUTES OF BATTERY LIFE - HP Fast Charge restores battery from 0 to 50% in approximately 45 minutes
- AMD RADEON 840M GRAPHICS - Built in for thrilling gaming performance, high resolution display support and hardware accelerated encoding with or without a discrete graphics card
- STORAGE AND MEMORY - 512 GB PCIe Gen4 NVMe M.2 SSD offers fast speed and efficient storage; and 16 GB DDR5 RAM memory boosts performance with higher bandwidth
| Your goal | What to verify |
|---|---|
| Use Recall or another Copilot+ feature | Exact Copilot+ PC model, supported NPU, Windows build, drivers, language and feature rollout. |
| Use Studio Effects | Windows 11 22H2 or newer, supported NPU, enabled manufacturer driver and compatible camera or microphone. |
| Run a downloaded language model | Model variant, available RAM and storage, CPU/GPU/NPU compatibility and a completed download. |
| Build an ONNX application | Windows ML runtime, packaged model, execution provider and a fallback plan for unsupported hardware. |
Microsoft has not published a controlled comparison establishing universal speed, battery or latency gains across Windows PCs. Those outcomes vary with the model, workload, drivers and hardware.
How to tell whether a task is truly offline
- Identify the exact feature. “Windows AI” is too broad; name Recall, a Click to Do action, Studio Effects, Foundry Local or an app’s own model.
- Check the processing statement. Look for documentation saying where inference occurs, not merely a marketing label such as AI PC.
- Separate preparation from use. Determine whether the model, catalog or language pack must be downloaded first.
- Trace follow-up actions. A local result can be passed to search, translation, rewriting or another provider that requires internet.
- Test after disconnecting. Turn off Wi-Fi or unplug Ethernet only after setup, then repeat the specific operation. A failure may indicate a missing cache rather than cloud inference.
Common failure cases and fixes
The feature is not visible
Check the exact PC model, Windows version, NPU capability, language and manufacturer driver. Copilot+ features are hardware- and rollout-specific; an ordinary Windows 11 installation does not automatically include them.
A previously working model stops offline
Confirm that the model files remain cached and that the application has not switched to a cloud fallback. Reconnect to refresh catalog metadata or complete an update, then test again before disconnecting.
Click to Do asks for a connection
You have likely selected an action that uses an online provider. Choose a documented local Phi Silica text action instead, or keep the connection disabled if the content must not leave the PC.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Studio Effects disappeared after an update
Recheck Settings > Bluetooth & devices > Cameras, install the manufacturer’s enabled driver, verify Windows 11 22H2 or newer and confirm that the camera is supported.
Privacy and security boundaries
Local inference reduces the need to upload the particular input used by that operation, but it is not a blanket privacy guarantee for Windows or third-party applications. An app may still synchronize files, send telemetry, require account services or invoke a cloud provider. Evaluate the network behavior of the complete workflow, not only the model’s location.
For sensitive work, document which model is installed, where its files are stored, which actions are local, and which menus or providers are disabled. Then verify the behavior with the computer disconnected before handling real data.
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.
