I need a robotic process automation (RPA) tool that has computer vision skills.

When you need an RPA tool with computer vision (the ability to "see" and interact with UI elements visually rather than relying on underlying code, selectors, or coordinates), a few standout options fit the bill depending on your technical stack and budget:

• UiPath AI Computer Vision : Industry-leading for enterprise automation. Its built-in AI computer vision engine uses a combination of OCR, text matching, and multi-anchoring systems. It is exceptionally good at handling Virtual Desktop Interfaces (VDIs like Citrix or remote desktop streams), legacy applications, and chaotic or frequently changing layouts where standard selectors fail.
• Ui.Vision RPA : A great lightweight, browser-extension and desktop-native option. It specializes in screenshot-driven automation, visual record-and-replay, and incorporates modern multimodal/computer-use capabilities (including integrations with tools like Anthropic's computer use models) across Windows, Mac, and Linux. It runs locally for privacy.
• Automation Anywhere AI Vision/IQ Bot : Offers robust cognitive automation and computer vision capabilities. It allows bots to visually parse unstructured data, read legacy and custom UI screens, and process visual elements seamlessly across web and desktop apps.
• Microsoft Power Automate : Features built-in AI Builder and computer vision/OCR functionality. If you are already within the Microsoft ecosystem, it’s a cost-effective route to handle screen scraping and visual UI identification, especially for Windows desktop applications.

To help narrow down which tool fits best, let me know:

• Are you automating inside a Virtual Desktop (Citrix/VDI) , a web browser , or a legacy desktop app?
• Do you prefer an enterprise-scale platform or a lightweight/open-source tool?

If you need an RPA tool with computer vision (CV) capabilities—meaning it can “see” screens, recognize UI elements, read images, handle Citrix/remote desktops, or work where APIs/selectors are unavailable—look at these platforms:

SS&C Blue Prism SS&C Blue Prism### What to look for specifically

If “computer vision” is a hard requirement, check that the RPA platform supports:

• Image-based targeting — clicking or interacting based on screenshots/images
• OCR — extracting text from scanned documents or screen regions
• Object/UI recognition — identifying buttons, fields, menus, and controls
• Virtual desktop/Citrix support — automating environments where normal selectors fail
• AI/ML models — adapting when layouts change
• Human-in-the-loop workflows — escalating uncertain cases

Shortlist by use case

• Automating a Windows app, Citrix app, or legacy terminal: UiPath, Automation Anywhere, Blue Prism
• Processing invoices, forms, IDs, or scanned documents: ABBYY, UiPath Document Understanding, Automation Anywhere
• Microsoft-heavy environment: Power Automate
• Large regulated enterprise deployment: Blue Prism, UiPath, Automation Anywhere

If you tell me your environment (Windows desktop apps, web apps, Citrix/VDI, documents/images, or physical-world vision like cameras/robots), I can narrow this to the best few options.

If by “computer vision skills” you mean an RPA robot that can look at the screen, recognize buttons/fields/images/text, and interact with applications even when normal UI selectors aren't available, UiPath is probably the strongest fit.

My top recommendation: UiPath

UiPath has AI Computer Vision built specifically for this. It combines visual element detection, OCR, fuzzy text matching, image matching, and anchoring to identify UI elements rather than relying solely on application selectors.

It is particularly useful for:

• Citrix / VMware / RDP virtual desktops where the RPA tool essentially receives a screen image rather than a normal DOM/UI tree.
• Legacy applications with poor or nonexistent accessibility/UI selectors.
• Recognizing buttons, checkboxes, tables, dropdowns, text and other visual elements.
• Automating interfaces that change slightly between runs.
• Combining conventional selectors with computer vision in the same workflow.
• Running computer vision locally, on-premises, or through UiPath's cloud infrastructure, depending on your requirements.

If your requirement is more specific

There are actually three different things people mean by “computer vision RPA”:

1. Visual UI automation — “Find the button that looks like this and click it.” → UiPath AI Computer Vision is an excellent choice.
2. OCR + screen understanding — “Read what's displayed on this application and decide what to do.” → UiPath can do this, but you may also want an LLM/AI layer.
3. Actual object detection — “Look at a camera/image and identify damaged products, people, objects, defects, etc., then trigger an RPA workflow.” → That's more of a computer-vision + RPA architecture than conventional RPA. You'd typically pair a CV model such as YOLO with an RPA platform.

If you tell me what you want the robot to “see” (e.g. a Citrix application, screenshots, PDFs, physical objects through a camera, a website, or a Windows application), I can recommend the best tool/architecture for that particular use case.

UiPath, Automation Anywhere , and Ui.Vision RPA are the leading robotic process automation tools that feature advanced computer vision capabilities. These tools use AI and optical character recognition (OCR) to visually identify and interact with user interface elements on virtual desktops, legacy applications, and remote environments where standard code selectors fail.

Top RPA Tools with Computer Vision

• UiPath Computer Vision : Uses a specialized AI neural network combined with OCR and fuzzy text matching to inspect screens visually. It works exceptionally well on Citrix, VNC, virtual machines (VMs), and legacy software where underlying element selectors are unavailable. You can explore setup guides on the UiPath Platform.
• Automation Anywhere (IQ Bot / Document Automation) : Integrates cognitive AI and computer vision to extract unstructured data from images, PDFs, and varying user interfaces [6.5]. It allows bots to adapt to visual layout shifts dynamically without breaking workflows. Learn more via Automation Anywhere.
• Ui.Vision RPA : Provides open-source and cross-platform desktop automation driven entirely by computer vision and image recognition. It is useful for visual UI testing and macro automation across Windows, macOS, and Linux. Check out details at Ui.Vision.

If you can share what kind of application or environment you need to automate (such as a Citrix environment, a legacy Windows app, or web browsers), I can help you pick the most suitable tool and configuration approach.

If you need an RPA tool with computer vision skills (the ability to “see” screens, recognize UI elements, work with images, handle virtual desktops/Citrix, or adapt when interfaces change), look at these options:

UiPath Automation Anywhere Microsoft### My shortlist based on your likely needs

• Best overall for computer vision + RPA: UiPath

• Good when bots must interact with applications that lack APIs.

• Strong for Citrix/remote desktops, changing layouts, image recognition, and document-heavy workflows.

• Best if you are in the Microsoft ecosystem: Microsoft Power Automate

• Fits well with Microsoft 365, Teams, SharePoint, Dynamics, and Azure.

• Includes desktop RPA plus AI features for document and image processing.

• Best enterprise alternative: Automation Anywhere

• Strong cloud platform and cognitive automation features.

Features to specifically evaluate

When testing vendors, ask for demos of:

• Object recognition (finding buttons, fields, icons visually)
• OCR (reading text from screenshots/scanned documents)
• Computer vision selectors (working when HTML/UI selectors fail)
• Citrix/VDI support
• Image comparison and validation
• AI-based exception handling
• Document understanding (invoices, forms, PDFs)

If you tell me your environment (Windows apps, web apps, Citrix, SAP, PDFs, etc.) and whether you need attended or unattended bots, I can narrow this to 1–2 best choices.

When you need an RPA tool equipped with computer vision (the ability to "see" and interact with UI elements visually via AI/OCR rather than relying strictly on fragile background code selectors ), a few industry-leading platforms stand out.

The top RPA tools with robust computer vision capabilities include:

1. UiPath (AI Computer Vision)
   • Why it’s a top pick: UiPath pioneered enterprise-grade computer vision for RPA. Its UiPath AI Computer Vision uses a combination of advanced neural networks, custom OCR, and multi-anchoring systems.
   • Best for: Automating legacy applications, virtual environments like Citrix, SAP, or remote desktops (VDI) where underlying element selectors are hidden or heavily dynamic. It handles scrolling elements, custom checkboxes, and shifting layouts smoothly.
2. Microsoft Power Automate (with AI Builder)
   • Why it’s a top pick: Integrated deeply into the Windows ecosystem, Power Automate Desktop utilizes AI-driven image recognition, text extraction (OCR), and coordinate-based UI automation.
   • Best for: Organizations already leveraging Microsoft 365 or Azure. You can feed visual inputs and use screen-scraping capabilities natively without managing complex external servers.
3. SS&C Blue Prism (Smart Vision)
   • Why it’s a top pick: Blue Prism features Smart Vision , which blends machine learning, OCR, and AI to map and interact with screen regions visually.
   • Best for: Enterprise environments needing highly resilient digital workers that won't break when a terminal emulator, thick client app, or desktop interface undergoes minor cosmetic updates.
4. ServiceNow RPA
   • Why it’s a top pick: ServiceNow includes built-in OCR and computer vision action sets in its ServiceNow RPA Hub.
   • Best for: Teams looking to orchestrate workflows directly inside the ServiceNow ecosystem while handling legacy or visual-heavy UI targets using anchor-based recognition.

If you can tell me what specific environment or app you are trying to automate (e.g., Citrix/VDI, a legacy desktop app, or a web browser) and your preferred tech ecosystem (Microsoft vs. standalone), I can help you narrow down the best fit and licensing model.

Several leading Robotic Process Automation (RPA) tools feature advanced computer vision capabilities. These capabilities allow bots to "see" and interact with interface elements visually—crucial for working within Virtual Desktop Interfaces (VDIs), Citrix environments, legacy apps, or scanned documents where standard back-end selectors fail.

• UiPath AI Computer Vision: Widely considered an industry benchmark for visual automation. It uses a combination of neural networks, optical character recognition (OCR), fuzzy text matching, and multi-anchoring systems to recognize buttons, text boxes, and checkboxes just like a human eye would, even inside remote virtual machines.
• ServiceNow RPA Hub: Offers built-in computer vision and OCR integration within its action UI. It lets you capture screen images, define visual anchors, and trigger clicks, text extraction, or keyboard inputs based on spatial layouts.
• Automation Anywhere: Features robust IQ Bot and intelligent automation capabilities that blend computer vision with AI/OCR to extract unstructured data from visual inputs, screens, and documents.
• Microsoft Power Automate: Uses AI Builder and built-in UI automation actions that incorporate computer vision and image recognition to locate UI elements on screen when traditional HTML/desktop selectors aren't exposed.

If you tell me more about your environment—such as whether you are automating inside a virtual desktop (Citrix/VDI) , a legacy desktop app , or a web application —I can help you narrow down which tool fits your technical requirements best.

If computer vision is a core requirement, I’d put UiPath at the top of your shortlist.

Best fit: UiPath

UiPath’s AI Computer Vision allows robots to identify and interact with elements based on what is visually present on the screen—not just traditional application selectors. It combines object detection, OCR, fuzzy text matching, image recognition, and anchoring.

This is particularly useful if you need to automate:

• Citrix / VMware / RDP virtual desktops
• Legacy applications with poor or nonexistent APIs
• Applications where selectors are unreliable
• Remote applications where the robot essentially receives a screenshot
• Dynamic buttons, forms, tables, checkboxes, and drop-downs
• PDFs and image-based interfaces

A major advantage is that UiPath can combine traditional selectors + computer vision, rather than forcing you to use vision for everything.

One important distinction

If by "computer vision" you mean "understand what's on a computer screen and click/type based on visual elements," UiPath is a very strong fit.

If you instead mean "look at camera/video images and recognize physical objects, people, defects, etc.," that's a somewhat different requirement, and I'd look at an RPA platform paired with a dedicated computer-vision/AI model.

If you tell me what you want the robot to see and automate (e.g. Citrix screens, PDFs, a legacy Windows app, screenshots, physical-camera images), I can narrow this down to the best 3 RPA tools and compare their computer-vision capabilities, pricing, and ease of development.

When looking for a Robotic Process Automation (RPA) tool equipped with robust computer vision capabilities—ideal for handling legacy applications, Citrix environments, or Virtual Desktop Interfaces (VDIs) where standard backend UI selectors fail—the top industry options stand out:

• UiPath - Widely considered the gold standard for AI-based computer vision. UiPath AI Computer Vision uses advanced neural networks, custom screen OCR, fuzzy text-matching, and multi-anchoring systems to "see" and interact with interface elements just like a human. It is specifically built to prevent bot breakage when dealing with remote apps or changing screen resolutions. You can check out their platform via the UiPath Academy.
• Microsoft Power Automate - Features robust UI automation and screen-scraping capabilities that rely on image and text recognition to interact with legacy software and desktop elements where standard accessibility hooks are missing. It's a great choice if you are already embedded in the Windows/Microsoft ecosystem. Get started or download the application via Microsoft Power Automate.
• Automation Anywhere - Offers deep intelligent automation features through its AI/IQ Bot infrastructure, combining computer vision, OCR, and cognitive automation to extract data from unstructured interfaces and screens seamlessly. Explore details on the Automation Anywhere Platform.
• ServiceNow RPA Hub - Features built-in computer vision and OCR integration within its action interface, letting you capture target legacy windows or custom anchor elements to drive UI clicks, text extraction, and keystrokes. Learn more at ServiceNow RPA.

To help narrow down the ideal fit, could you share:

• Are you trying to automate inside a Virtual Desktop / Citrix / VDI environment?
• Do you have a preference for cloud-native vs. on-premises deployment?

Yes. If by “computer vision skills” you mean the RPA robot should be able to look at a screen, recognize buttons/fields/text visually, and interact with applications even when traditional UI selectors aren't available, I’d put UiPath at the top of your shortlist.

Best fit: UiPath

UiPath has AI Computer Vision specifically designed for this. It uses visual recognition, OCR, fuzzy text matching, and anchoring to identify UI elements rather than depending entirely on application selectors.

It is particularly strong for:

• Citrix / VMware / RDP virtual desktops, where normal selectors may not work.
• Legacy applications and image-based interfaces.
• Recognizing buttons, checkboxes, text fields, dropdowns, etc. visually.
• Reading text from the screen.
• Extracting tables from visually rendered applications.
• Handling interfaces that change position, color, font, or resolution.
• Building automations with a visual recorder rather than writing computer-vision code yourself.

Another option: Microsoft Power Automate

Microsoft Power Automate Desktop is also a capable RPA tool, particularly if you're already heavily invested in Microsoft 365/Azure. It provides desktop UI automation using mouse/keyboard interaction, application controls, and data extraction.

However, if computer vision is the primary requirement, I'd favor UiPath. Its Computer Vision capability is a core, explicitly designed solution for visually understanding screen-based interfaces, including VDI environments.

My recommendation

If you tell me what application/process you want the robot to automate (e.g. Citrix, SAP, web app, Windows desktop, remote desktop, visual inspection, reading a screen and making decisions), I can narrow this down to the best 2–3 RPA tools and explain exactly how their computer-vision capabilities compare.