
Telerik AI Coding Assistants
A family of Model Context Protocol servers from Progress that feed Telerik and Kendo UI knowledge straight into the AI coding assistant a developer already uses, so generated .NET and JavaScript component code needs less rewriting.
What is Telerik AI Coding Assistants?
The Progress Telerik and Progress Kendo UI AI Coding Assistants are MCP servers that supply proprietary context to whichever AI-enabled IDE a developer already works in. They do not replace GitHub Copilot, Cursor or Claude Code; they sit alongside them and correct a specific weakness. Left to themselves, general-purpose models guess component names and properties, produce wrong imports and reach for outdated patterns, so the developer spends the saved time fixing bindings and examples instead.
Each assistant is built from the API surface, documentation, source code and examples of one Telerik or Kendo UI library. Progress lists twelve of them, spanning Blazor, React, Angular and jQuery on the web side, and WPF, WinForms, .NET MAUI, ASP.NET Core, ASP.NET MVC, ASP.NET AJAX, Reporting and Document Processing beyond it. They install as NPM or NuGet packages, such as @progress/telerik-blazor-mcp, and are declared to the editor through a standard MCP configuration file. An On-Premises deployment is documented for teams that need it.
The same product page carries a second, higher tier. Where the assistant helps at the level of an individual component, the Telerik Agentic UI Generator turns a prompt into a complete page: it picks components, sets layout and theme, binds data and writes the code into the project, and it includes everything the assistant does. It is available for Blazor, Angular, React and ASP.NET Core, and ships with the DevCraft bundles, KendoReact, Kendo UI and Telerik UI for Blazor.
Progress reports customer-observed gains of up to 30% extra efficiency for the assistant and up to 50% extra productivity for the generator, alongside a 45% cut in token consumption on selected products. Accessibility is treated during generation rather than afterwards, with WCAG 2.2 Level AA guidance, ARIA usage, keyboard navigation and semantic markup. Developers keep their own model, cloud-hosted or local, and Progress states that prompts are never used to train Telerik AI models. Access is bundled: any active subscription includes the MCP servers, a 30-day trial covers them, and perpetual licences are excluded outright.
What it does
- Generate Telerik and Kendo UI component code directly inside your IDE
- Configure and fine-tune an existing component through prompting
- Generate complete styled, accessible pages with the Agentic UI Generator
- Produce dummy data and wire up data binding for prototyping
- Get step-by-step explanations drawn from the official documentation
- Troubleshoot straightforward component problems without leaving the editor
- Convert standard WinForms controls and designer elements into Telerik equivalents
When to use Telerik AI Coding Assistants / When not to
A quick filter to help you decide if Telerik AI Coding Assistants is the right fit.
When to use Telerik AI Coding Assistants
- Development teams that already hold a Telerik or Kendo UI subscription and want more from it
- Developers who use GitHub Copilot, Cursor or Claude Code and keep correcting component code by hand
- .NET and JavaScript engineers working across Blazor, React, Angular, jQuery, WPF or WinForms
- Enterprise teams that must ship interfaces meeting WCAG 2.2 Level AA from the outset
- Organisations with a mandated LLM, whether cloud-hosted or deployed on their own infrastructure
When not to use Telerik AI Coding Assistants
- Teams with no Telerik or Kendo UI licence, since an active subscription is a hard prerequisite
- Holders of a perpetual licence, who are explicitly excluded from the assistant
- Developers looking for AI help with styling and theming, which is handed off to ThemeBuilder
- Anyone without an IDE and an MCP-capable client already configured
- Projects built outside the Telerik and Kendo UI ecosystem, where the added context is worthless
How to use Telerik AI Coding Assistants
A typical end-to-end flow, from setup to results.
- Confirm you hold an active Telerik or Kendo UI subscription, or start a 30-day trial
- Create a Telerik account, which the assistant uses to check your licence entitlement
- Install the component library for your framework into the application
- Pick the matching MCP server package and install it from NPM or NuGet
- Declare the server in your MCP client, either in a workspace mcp.json or globally
- Restart the editor and confirm the Telerik tools appear in the client's tool list
- Select the model you want to drive it, cloud-hosted or deployed locally
- Prompt in Ask mode for explanations, or Agent mode for code the assistant writes into the project
- Start a fresh chat session when you switch task or file, to keep the context clean
- Review every generated diff before committing it, as Progress explicitly recommends
Pros & Cons
Pros
- Context drawn from the vendor's own API, documentation and source code, not from public guesswork
- Slots into the IDE and assistant already in use, with no change of workflow
- Included at no extra cost in any active subscription, and covered by the free trial
- The customer picks the model, which suits organisations with an approved or local LLM
- Explicit commitment that prompts are not used to train Telerik AI models
- Accessibility handled during generation rather than retrofitted afterwards
- Backed by an established vendor holding first-party ISO/IEC 27001 and SOC 2 attestations
Cons
- Useless without a Telerik or Kendo UI subscription, which is a hard prerequisite
- Perpetual licence holders are excluded from the assistant altogether
- No price of its own, so the real cost depends on which library you buy
- A shared fair-usage quota covers every Telerik AI tool on the same account
- The Agentic UI Generator reaches only part of the range, not all twelve frameworks
- Value is confined to the Telerik and Kendo UI ecosystem and nothing beyond it
- Progress itself advises against exposing proprietary or production data to the tool
Pricing & Plans
There is no permanent free plan, and no price attaches to the AI Coding Assistants themselves: they are bundled into any active Telerik or Kendo UI subscription rather than sold separately. A 30-day free trial includes them, granted automatically when a product trial begins, while perpetual licences are explicitly excluded. Because the entitlement rides on the library purchased, the entry point varies considerably. The cheapest subscription carrying an assistant is Telerik Reporting at 499 USD per developer per year on the Lite support tier; Kendo UI starts at 799 USD, KendoReact at 649 USD, and the DevCraft bundles run from 1,149 USD to 1,649 USD. The 899 USD figure shown on the product page is the Priority support tier for Kendo UI, not that product's entry price. No single amount fairly represents the tool, since each subscription unlocks a different assistant.
- 30 days
- includes the AI Coding Assistant
- granted automatically with a product trial
- 499 / 599 / 999 USD per developer per year
- 649 / 749 / 1
- 199 USD per developer per year
- 749 / 849 / 1
- 249 USD per developer per year
- 799 / 899 / 1
- 299 USD per developer per year
- MCP servers included
- 1
- 149 USD per developer per year
- 1
- 299 USD per developer per year
- 1
- 649 USD per developer per year
- Lite (72h response
- 10 incidents)
- Priority (24h
- unlimited)
- Ultimate (phone support and escalation)
Data, GDPR & hosting
A consolidated view of how Telerik AI Coding Assistants handles your data.
GDPR overview
Progress publishes a Data Processing Addendum, a 27-page document that addresses Article 28 GDPR and also covers the UK GDPR, Brazil's LGPD, the CCPA and Swiss law. It incorporates the European Commission's Standard Contractual Clauses, and names the Netherlands as the default supervisory authority where no other is identified. Data subject requests run through a dedicated form, and privacy queries reach the PSC Data Protection Office in Burlington, Massachusetts. Progress states plainly that it is a US-headquartered company and that covered information is hosted and stored in the United States. Two gaps are worth naming: Progress nowhere claims GDPR compliance in so many words, and no Article 27 EU representative is designated in any page examined.
Who owns the data?
Progress states that the assistant has no access to your workspace or application code. It cautions, however, that the LLM you choose generates the parameters sent to the MCP server, and those parameters may carry fragments of your code to that model. Prompts and generated context are anonymised and are not tied to your Telerik account; Progress keeps them for statistical and troubleshooting purposes, and keeps usage metrics to police your licence quota. Under the Progress Data Processing Addendum, the customer remains controller of any personal data and Progress acts as processor. Progress explicitly refers users to the terms and privacy policy of whichever AI model and client they select.
Reuse rights
The code the assistant helps produce is yours to keep and ship: Progress positions the output as clean code you retain, with no reuse permission to request. Two limits are documented rather than legal. Progress tells users to double-check every solution before applying it to a project, and to avoid exposing proprietary or production data to any AI-enabled tool. On its own side, Progress commits that prompts are not used to train Telerik AI models, and that the assistant never produces the answer itself, only the context handed to the model you chose. Whatever that third-party model does with the request is governed by its own terms.
Data retention & training
Hosting summary
Progress states plainly that it is a US-headquartered company and that information covered by its privacy policy is hosted and stored in the United States, a jurisdiction it acknowledges does not offer data protection comparable to the European Union. Using the service is presented as agreement to that storage and processing. Transfers out of the EU are framed by the European Commission's Standard Contractual Clauses, incorporated into the Progress Data Processing Addendum, which names the Netherlands as the default supervisory authority where no other is identified. No hosting country other than the United States is named anywhere in the pages examined, and no regional data residency option is offered for this product. One meaningful exception applies: Progress documents an On-Premises deployment of the AI Coding Assistant, in which case hosting becomes the customer's own. Note too that the model answering your prompts is one you select, and its hosting is governed by that provider rather than by Progress.
Things to keep in mind
Risks and trade-offs to weigh before adopting Telerik AI Coding Assistants.
- The vendor itself tells you to double-check every generated solution; treating the output as correct by default is the main risk to code quality
- Progress advises against exposing proprietary or production data to AI-enabled tools, yet the LLM may place fragments of your code in the MCP request parameters
- The product page advertises 899 USD while the vendor's own pricing grid starts at 799 USD for Kendo UI and 499 USD for Reporting; the page's structured data repeats the higher figure
- The legal entity behind the tool, Progress Software Corporation, was constituted in Delaware on 8 May 2015, while the company advertises forty years of activity and the product itself dates from 2025
- Leaning on generated component code can erode a developer's own grasp of the library, which matters when the assistant is wrong or unavailable
- The product page still carries unedited template placeholders in a video block, and the announcement blog describes GitHub Copilot extensions discontinued in November 2025
- Detailed privacy and quota terms were read on the Kendo UI for jQuery documentation; the other eleven assistants may differ in the specifics
Setup & Integrations
Technical difficulty
Squarely a developer task, and moderate for that audience. You need a configured IDE, an MCP-capable client, the component library installed and an active licence before starting. Installation means pulling an NPM or NuGet package and hand-editing a JSON configuration file; there is no graphical installer. Anyone comfortable with package managers and editor configuration will finish in minutes, guided by per-framework documentation and an On-Premises guide. Non-developers cannot use this product at all.
Deployment
Integrations
Supported languages
Behind Telerik AI Coding Assistants
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What is the difference between the AI Coding Assistant and the Agentic UI Generator?
Which AI model does it use?
Are my prompts used to train Telerik models?
Can the assistant read my source code?
Is it included in my licence?
How do I install it?
Which frameworks are covered?
Can I run it on our own infrastructure?
Is there a usage limit?
Is there a minimum age?
Should you pick Telerik AI Coding Assistants?
The Telerik and Kendo UI AI Coding Assistants are best understood as a correction layer rather than a coding assistant in their own right. They assume a developer already has an IDE, an AI assistant and a Telerik or Kendo UI subscription, and they address the narrow but real problem of a general-purpose model inventing component APIs. Within that scope the proposition is coherent: the context comes from the vendor's own documentation and source code, twelve frameworks are covered, and nothing about the existing workflow has to change.
Two commitments carry unusual weight for enterprise buyers. Progress lets the customer keep their own model, cloud-hosted or local, and states without conditions that prompts are not used to train Telerik AI models. An On-Premises deployment and first-party ISO/IEC 27001 and SOC 2 attestations reinforce that posture, though those attestations cover Progress as a company rather than this product specifically.
The commercial framing is the part that deserves care. The assistants have no price of their own and no standalone existence; they are an argument for renewing a component library subscription. Entry costs range from 499 to 1,649 USD per developer per year depending on which library you buy, each unlocking a different assistant, and perpetual licence holders get nothing at all. The 899 USD advertised on the product page is a middle support tier, not a floor. Prospective buyers should also note that the productivity gains quoted, up to 30% and up to 50%, are attributed to customer reports rather than to published measurement.
For a team already invested in Telerik or Kendo UI, this is a well-judged addition at no marginal cost. For anyone else, it is not a reason to adopt the platform.
- Choosing a selection results in a full page refresh.
- Opens in a new window.