JLanguageTool is running slow

Question

I am trying to implement JLanguageTool in combination with JavaFX and RichtextFX to get myself a GUI which spellchecks the given input in a Codearea and highlights misspelled words. My problem is, that even for small sentences (6 words) with a little amount of misspells (3 misspells) my program takes 10-12 seconds to execute(only the spellchecking part). I've already tried optimizing my program by deactivating rules that are not neccessary for me. Here's my code:

import org.languagetool.JLanguageTool;
import org.languagetool.language.GermanyGerman;
import org.languagetool.rules.Category;
import org.languagetool.rules.CategoryId;
import org.languagetool.rules.RuleMatch;

import java.io.IOException;
import java.time.Duration;
import java.time.Instant;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;

public class LanguageToolTest {

    private static JLanguageTool tool = new JLanguageTool(new GermanyGerman());

    public static void prepareRule(){
        Map <CategoryId, Category> akku = tool.getCategories();
        for (var entry : akku.entrySet()){
            if (!entry.getKey().toString().contains("EMPFOHLENE_RECHTSCHREIBUNG")
            && !entry.getKey().toString().contains("CASING")
            && !entry.getKey().toString().contains("TYPOS")
            && !entry.getKey().toString().contains("COMPOUNDING")
            && !entry.getKey().toString().contains("PUNCTUATION")
            && !entry.getKey().toString().contains("CONFUSED_WORDS")
            && !entry.getKey().toString().contains("SEMANTICS")){
                tool.disableCategory(entry.getKey());
            }
        }
    }

    public static void main(String[] args) {
        prepareRule();

        Instant start = Instant.now();
        Map<Integer, List<String>> test = new HashMap<>();
        int index = 0;
        try {
            List<RuleMatch> matches = tool.check("Heilike makrele hier ist echter Tecst");
            for (RuleMatch match : matches){
                List<String> temp = new ArrayList<>();
                index = match.getFromPos();
                int end = match.getToPos();
                String idk = "Heilike makrele hier ist echter Tecst".substring(index, end);
                temp.add(idk);
                temp.addAll(match.getSuggestedReplacements());
                test.put(match.getFromPos(), temp);
            }
            Instant stop = Instant.now();
            Duration timeElapsed = Duration.between(start, stop);
            System.out.println(test.toString());
            System.out.println("Hat " + timeElapsed.toMillis() + " Millisekunden gedauert.");
        } catch (IOException e) {
            e.printStackTrace();
        }
    }
}

And the output:

{0=[Heilike, Heilige, Heiligen, Heidecke, Heilig, Heinicke, Heiliger, Heiliges, Heiligem, Beilage, Heiligt, Heike, Heiligte, Fällige, Herlocke, Heutige, Wellige, Heftige, Eilige, Häufige, Seidige], 32=[Tecst, Eckt, Deckt, Test, Täks, Deckst, Erst, Teast, Lässt, Des, Meist, Fest, Heißt, Jetzt, Setzt, Weist, Recht, Text, West, Wächst, Geist], 8=[makrele, Makrele, makele, mäkele, sakrale, magere, mangele, Makrelen, Nägele, Häcksel, Häckseln, angele, dackele, fackele, fächele, hagele, häckseln, häkele, lächele, magerere, magerte]}
Hat 12483 Millisekunden gedauert.

To explain the Map: my Key stores the start index of the found error for further processing and the List value stores a list of the misspelled word at index 0 and recommended replacements on the following indecies.

In my actual code the Map gets returned and processed for the highlighting part.

As far as I understood the documentation of JLanguageTool it's not thread-safe therefore I can't multithread it.

Any suggestions on how to improve my code/optimize certain calls to reduce the performance impact?