> For the complete documentation index, see [llms.txt](https://docs.mindee.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.mindee.com/integrations/client-libraries-sdk.md).
# Client Libraries / SDKs
By using the client libraries you'll be able to integrate faster and lower your maintenance costs.
Some useful tools are also provided, for example PDF processing and image compression.
You can use the client libraries to make API requests following the polling and webhook patterns.
All our client libraries are open-source (MIT license) and hosted on [GitHub](https://github.com/mindee).
Supported languages/frameworks: **Python**, **Node.js** (JS/TS), **PHP**, **Ruby**, **Java**, **.NET** (C#).
## Installation Instructions
{% tabs %}
{% tab title="Python" %}
Requires Python ≥ 3.9. Python ≥ 3.11 is recommended.
Simply install the [PyPi package](https://pypi.org/project/mindee/) using `pip`:
```sh
pip install -U mindee~=5.1
```
{% endtab %}
{% tab title="Node.js" %}
Requires Node.js ≥ 20.1. Node.js ≥ 22 is recommended.
Simply install the [NPM package](https://www.npmjs.com/package/mindee):
```sh
npm install mindee@^5.5.0
```
{% endtab %}
{% tab title="PHP" %}
Requires PHP ≥ 8.1. PHP ≥ 8.3 is recommended.
Simply install the [Packagist package](https://packagist.org/packages/mindee/mindee) using [composer](https://getcomposer.org/):
```sh
php composer.phar require "mindee/mindee:>=3.0"
```
{% endtab %}
{% tab title="Ruby" %}
Requires Ruby ≥ 3.2.
Simply install the [gem](https://rubygems.org/gems/mindee) using:
```shell
gem install mindee -v '~> 5.2'
```
{% endtab %}
{% tab title="Java" %}
Requires Java ≥ 11. Java ≥ 17 is recommended.
Group ID: `com.mindee.sdk`\
Artifact ID: `mindee-api-java`\
Version: `5.2.0` or greater
There are various installation methods, Maven, Gradle, etc:
[Installation Details](https://central.sonatype.com/artifact/com.mindee.sdk/mindee-api-java)
{% endtab %}
{% tab title=".NET" %}
.NET ≥ 8.0 is recommended.
Simply install the [NuGet package](https://www.nuget.org/packages/Mindee) using `dotnet add`:
```sh
dotnet add package Mindee --version 4.4
```
{% endtab %}
{% endtabs %}
Don't see support for your favorite language or framework? [Make a feature request!](https://feedback.mindee.com/?b=682f69c9e2404756e7e68d1c)
## Usage Details
Overall, the steps to using the Mindee service are:
1. [Client Configuration](/integrations/client-libraries-sdk/configure-the-client.md)
1. Initialize the Mindee client.
2. Set inference parameters, in particular the model ID to use.
2. [Load and Adjust a File](/integrations/client-libraries-sdk/load-and-adjust-a-file.md)
1. Load a file from various supported sources: path, bytes, etc.
2. *Optional*: adjust the source file before sending.
3. [Send a File or URL](/integrations/client-libraries-sdk/send-a-file-or-url.md)
1. Send the file or an URL with the proper parameters.
4. [Response Processing](/integrations/client-libraries-sdk/process-the-response.md)
1. Optional: load from a webhook.
2. Optional: access document metadata
5. [Extraction Result](/extraction-models/sdk-integration/extraction-result.md)
1. Handle the field values extracted from the document
2. Optional: access field metadata (polygons, confidence score)
## Frequently Asked Questions
Can I send requests in parallel?
Yes. All clients can be used to send requests in parallel.
The exact implementation is left to the user:
* For Node.js, you'll want to use asynchronous processing.
* For all others, you'll want to use threads or processes.
Can I use the v1 and v2 APIs together?
Yes. Each client library has support for both v1 and v2 APIs.
You'll need to make a separate instance of the client classes:
* .NET and Java, use `MindeeClient` and `MindeeClientV2`
* all others, use `Client` and `ClientV2`
The code to make requests and to process results is **very different** between v1 and v2.
We highly recommend having different files (or even modules) for handling each API version.
Can I send only a specific page of a multi-page PDF?
Yes. All libraries have support for cutting/extracting PDF pages.
For more information, consult: [Load and Adjust a File](/integrations/client-libraries-sdk/load-and-adjust-a-file.md#manipulate-pdf-pages).
How to stop PDFs with too many pages from being sent?
**Do not use file size**, a text PDF with 200 pages can be smaller than a single photo.
Much more reliable to count the actual number of pages in the PDF document.
Use the built-in file metadata methods and properties to easily add business rules based on the number of pages (among other data).
For more information, consult: [Load and Adjust a File](/integrations/client-libraries-sdk/load-and-adjust-a-file.md#source-file-metadata).
I'm using a Supabase edge function, should I use the API directly?
We recommend using the Mindee [Node.js client library](https://github.com/mindee/mindee-api-nodejs) in Supabase.
You can install it in your edge function(s) using `npm`.
Which library features are officially supported?
Anything documented here is officially supported and is considered stable for production use.
Anything in a library that is not documented here, is **not** officially supported and subject to change or removal.
There is a bug with the SDK, how can I get help?
If you are encountering a persistant issue, you should try first:
1. Asking the Docs Assistant for help:
2. Asking your local agent for help by using one of our [SKILL files](/integrations/ai-coding-assistants.md#skill-files)
After this, if you determine (or are told) there is a problem with the SDK itself, the best bet would be to open a bug report on the relevant SDK's GitHub repository:
* **Python**:
* **Node.js**:
* **PHP**:
* **Ruby**:
* **Java**:
* **.NET**:
Our SDK engineers will be automatically notified of the new issue.
**Note:** Pro and above plans can contact our support teams directly.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://docs.mindee.com/integrations/client-libraries-sdk.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.