FRENES

Glossary · MCP protocol

SDK

SDK stands for software development kit. It is a collection of reusable code, documentation, and sometimes tools that helps developers build applications for a service or platform, such as connecting your AI assistant to a language model without manually implementing every network exchange.

SDK: the explanation

An SDK packages common operations into functions that a developer can call. Instead of assembling every request to a service by hand, your application passes information to the kit. The SDK prepares the request and makes the response easier to use. Depending on its design, it may also handle time limits, certain errors, or retries. Developers still need to configure access and understand which information the application sends, receives, and stores during these exchanges.

SDKs help developers work with existing services more efficiently. In an AI application, a kit might simplify sending messages to a model, receiving an answer progressively, or connecting a document service. You can benefit from this work without installing anything yourself because the SDK is part of the application. A service can offer several SDKs for different programming languages, all communicating through the same API, which defines the available requests and responses for that service.

An SDK is different from an API. The API defines how software can interact with a service, while the SDK provides convenient building blocks for that interaction. Coverage can vary, and a kit may not immediately support every newly available function. Developers therefore need to check its documentation and maintenance history. They should also distinguish kits maintained by the service provider from community projects, which may have different support arrangements and approaches to updates or compatibility.

A concrete example

You want meeting summaries to appear inside your organization’s notes application. A developer uses an SDK to send approved meeting notes to a model and retrieve a summary. They then add the interface, privacy rules, and error handling around that connection. The SDK reduces some integration work, but it does not decide which notes may leave your organization or whether the resulting summary is accurate enough to share.

Why it matters

An SDK affects how easily an AI tool can be integrated and maintained. If you expect to customize a product, check its supported programming languages, documentation, and update practices. A convenient kit can speed up development while making a later provider change more involved. Having an SDK does not establish that an application protects your information properly or that its underlying model produces dependable answers for your tasks.

Common misconceptions

  • An SDK is not always required to use an AI service: developers can communicate directly with its API, while ordinary users can access an application that handles those technical exchanges entirely behind the interface they see.
  • An official SDK does not automatically make an application secure: the surrounding software still needs appropriate permissions, protected credentials, careful data handling, and checks on model responses before those responses trigger consequential actions for you.

Related terms: mcp protocol

MCP (Model Context Protocol)
MCP, or Model Context Protocol, is a protocol for exchanges between an AI application and services that provide data or functions.
MCP server
An MCP server is a software component that exposes data or functions through the Model Context Protocol.
MCP client
An MCP client is the component of an application that communicates with a server using the Model Context Protocol.
API
An API, or application programming interface, defines how software can use another component’s functions or data.
webhook
A webhook is a mechanism that lets a service automatically send a notification to a web address when an event occurs.
structured output
Structured output is a response arranged according to a defined format, with expected fields and data types.
A2A protocol
The A2A protocol is an open set of rules that lets independent AI agents communicate and delegate tasks.

Where this site covers it

The term appears on 155 pages of this site. A few of them:

Go further

Frequently asked questions

Do I need programming skills?

Directly integrating an SDK generally requires programming in a supported language. If you are using a finished application, you normally do not need to know which SDK it uses or install that kit on your own device.

Is an SDK a connector?

A connector usually links services for a particular workflow. An SDK gives developers reusable components for building integrations. A connector may rely on an SDK internally, but users do not normally need to interact with that kit.

Why update an SDK?

Updates can fix problems, support changes to an API, or expose additional functions. They can also alter existing behavior. The team maintaining an application should check compatibility and relevant workflows before deploying an updated SDK.

Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score