Skip to content
View in the app

A better way to browse. Learn more.

LCPDFR.com

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

NPCI 2.1.0.0

(15 reviews)

2 Screenshots

NPCI 2.1 is LIVE!

 

NPCI brings you a new way of interacting with PEDS inside of LSPDFR and adds a whole new level of immersion to your gameplay experience!

 

ANNOUNCEMENT:

The new update (2.1) is OUT. This new version finally brings the Microsoft Speech Platform to NPCI, but with a twist! You can now CHOOSE between WSR or MSP in the INI file. You can also tweak the confidence level threshold for each engine in the INI file to find your sweet spot. Lower the value to loosen up the speech recognition or increase the value to tighten it up (you can do it in .01 increments). 

 

*** Because both engines (WSR & MSP) are intertwined in the new NPCI 2.1 update - YOU MUST have Microsoft Speech Platform installed to use NPCI 2.1, even if you don't intend on using the new Microsoft Speech Platform. If you do not have MSP installed, you can download it from the links below: *** 


Microsoft Speech Platform: https://www.microsoft.com/en-us/download/details.aspx?id=27225 
Language Packs:  https://www.microsoft.com/en-us/download/details.aspx?id=27224
-- For the speech platform install, download the larger 2.6 MB SpeechPlatformRuntime.msi, NOT the 2.2 MB one. 
-- For English users - download the "MSSpeech_SR_en-US_TELE.msi"

 

Here's the complete list of changes in NPCI 2.1:

Spoiler

• Support for Microsoft Speech Platform -- now anyone can use NPCI where MSP is supported!

• DUAL speech recognition option. Set the engine to use WSR or MSP in the NPCI.INI file.
• Speech recognition confidence threshold customization - modify the confidence threshold in the NPCI.INI file.

• Fixed Grammar Police PED check not working when asking for license on foot or during traffic stops (PR version - Huge thanks to Echooo for that one). 
• New blacklist option for all emergency personnel (default is enabled).
• New PED actions by voice (PED wait by patrol vehicle & PED go to crosshair location & more).
• New Player actions by voice.
• New PED automark feature for both PR & STP (auto enables PR/STP when you perform my frisk action).
• XML update that brings nested phrases into a single command.
• Controller support for NPCI voice input and for the new PedGoToCrosshair action.

 

For those of you who just want to add the new actions into your existing UserResponses xml file (so you don't overwrite any of your existing customization's), I included a text file with the new actions. Just copy and paste them into your existing xml file and you'll be good to go. Don't forget to take advantage of the new XML nested phrases feature!

 

Need support? Join the NPCI Discord and help grow the NPCI community - join us here: https://discord.gg/9RMrrvm5qH

Want to show some love for all my hard work on NPCI - join my Patreon here: https://www.patreon.com/cw/Doublehook

 

How to use the new PED go to crosshair location action:

Spoiler

First, customize the [Targeting] keybind to whatever you prefer in the NPCI.INI file. Then while in game, hold down the keybind you set for the [Targeting] crosshair, and you'll see a small white crosshair appear on your screen. Move that around to your desired location, and while holding down the targeting keybind, give the phrase to the PED to "wait over there" or "go over there" or whatever you customize this phrase with, and then let off the [Targeting] keybind, and the PED will comply. Your player character will automatically turn to the location you selected and point for the PED to go. There are 2 different actions for this new feature - one that commands the PED to go to your crosshair and then turn around and face you and the other commands the PED to go to your crosshair and face away. Great to use the face away one when you want PEDS to get up against a vehicle or wall for pat downs ("get against the vehicle" or "get against the wall"). It might take you one or 2 tries to get the hang of how it works, but once you do, you'll end up using these actions a ton!

 

 NEW XML Nested Phrases Update:

Spoiler

You can now nest phrases inside of a single command, to help better organize your phrases and responses and avoid all the duplication with variations of your phrases. An example of this is shown below:

 

    <Command>
        <Text>take care</Text>
        <Text>take it easy</Text>
        <Text>you have a good one</Text>
        <Text>have a good one</Text>
        <Text>have a good day</Text>
        <Text>you have a good day</Text>
        <Text>have a good evening</Text>
        <Text>you have a good evening</Text>
        <Text>have a great day</Text>
        <Text>have a great evening</Text>
        <Action>DismissNPC</Action>
        <Responses>
            <Response>See you later sir.</Response>
            <Response>See you officer.</Response>
            <Response>You too sir.</Response>
            <Response>You too officer.</Response>
            <Response>Will do sir.</Response>
            <Response>Will do officer.</Response>
            <Response>You do the same.</Response>
        </Responses>
    </Command>

 

NPCI v2.1 brings a ton of new features, such as PED actions by voice, Player actions by voicenew PED animations, PED idle animations, a built in animation viewer, the ability to cycle through 49 different walk styles (with the option to set a default walk style when loading up LSPDFR), and more! Interactions in NPCI 2.1 have never felt more alive!

 

There are 2 versions available in the download: "NPCI 2.1 for PR" & "NPCI 2.1 for STP". Make sure you install the correct one!

 

Installation:

Spoiler

First and foremost, please read through ALL of the provided documentation before you get started. There are so many new features in version 2.1, and I'd hate for you to miss out on everything that NPCI 2.1 has to offer. 

 

NPCI is now an LSPDFR plugin. Before you install the new version, be sure to remove the old version from the base plugins folder, or you'll encounter issues.

 

Drag and drop the "plugins" folder inside the zip for the version you're using (PR or STP) directly into your main Grand Theft Auto "root" folder, or simply drag and drop the individual plugin files (including the WindowsInput.dll) directly into your "plugins\LSPDFR" directory. Make sure the NPCIXML folder is also in the same "plugins\LSPDFR" directory as the dll.

 

\Grand Theft Auto\plugins\LSPDFR\

 

For users who have significantly customized their UserResponses.XML files, I included a text file in the NPCI 2.1 zip with the new actions that were added in this update, so you can simply copy and paste those new actions directly into your existing UserResponses.XML. 

 

Due to the new version and how the XML files have changed, I've re-provided BOTH the "Responses.xml" & "UserResponses.xml" files. Don't OVERWRITE your customized versions!! For those of you who have customized your xml files, you'll need to update them to the current format. All responses, regardless of the xml file, have a new "<Action> </Action>" tag. This tag is vital to NPCI 2.1, as it dictates what actions or animations will trigger when you give a voice command phrase. All of your custom phrases will need to have this new action tag added into your xml files (if you want the PEDS to respond with the new animation features). There is a README in the download that covers all of the new PED actions and PED animation category tags, so you'll know which ones to add to your custom responses for certain situations. 

 

XML Customization

Spoiler

Please refer to the "NPCIXML" folder inside the plugins directory to customize the questions/responses and the new "Action" commands. There are 2 XML files to customize these features. One is called "Responses.XML", which is the default XML file containing all of the original questions/responses from v1.0.3. The 2nd XML file is called "UserResponses.XML", which contains all of the action commands (along with some random questions/responses as well). You can use either one to customize your very own questions and responses for NPCI. Either XML file will also take the action commands. In addition to these XML files, you will find 3 new ones called GenericPedAnimations.xml, IdleAnimations.xml, & AnimViewer.xml. The provided documentation explains what these XML files are for and how you can customize them. I will be putting up some videos in the next couple of days walking you through the new XML files and how you can customize them. 

 

**Quality of life update: the questions/phrases can now have capital letters in there, so no more worries about them all having to be lowercase!

 

**** Due to in-game limitations, the max response length is 100 characters. Anything over 100 characters will be cutoff on the onscreen response. When customizing your responses, you must keep them to 100 characters or less (between the <Response> and </Response> brackets).

 

Demonstrations

These videos are using voice activation and NPCI 2.1 only. No menus, just straight up voice commands. I just added an early preview for NPCI 3.0 as well, so enjoy! 

 

Keep an eye out on my YT page (https://www.youtube.com/@doublehookmedia) for more videos! I'll be posting videos to show you how to modify your XML files, including the PED animations and idle animations. I'll also go over how to create new PED emotion categories and explain how to customize them to create your own PED animations.

 

What is NPCI?

Spoiler

NPCI stands for "Non Playable Character Interaction". NPCI allows you to talk to any PED inside of GTA 5 LSPDFR using just your voice and get text responses back onscreen. Like many others, I found myself avoiding the "question the ped" option inside of STP, because it took too long to generate the interaction dialog. Desperately wanting to include this immersive interaction in my gameplay, I set out to create a plugin for myself that would allow me to speak the same types of questions using just my voice. After 5 long months of testing and tweaking, it ended up working out better than I could have ever hoped for. Now I'm sharing that same experience with all of you!

 

Suggested "must have" LSPDFR plugins & addons to use alongside NPCI

Policing Redefined & Grammar Police. While NPCI works the same for both PR & STP when it comes to the available features, NPCI runs much smoother with the better coded Policing Redefined, so if you want the best NPCI experience, I strongly suggest you go with PR!

 

Tutorials

Spoiler

The new NPCI.ini file comes with several options for keybind customization's. You can customize your voice input (now with a modifier option & controller support), a new RP type mouth movement that you can use to talk to PEDS when not asking them questions from the XML file, and many more. ** A neat trick you can implement -- if you have a way of stacking keybinds on your mouse, you can use both my new roleplay (RP) manual mouth movement keybind option, along with Grammar Police, to make your mouth move when you talk to dispatch! The stack order does matter, so my RP keybind would have to come before GP, otherwise, GP's keybind cancels the mouth movement ** Don't forget to add a custom walkstyle name in the INI file to load when LSPDFR starts up! All of the keybind options come with a modifier option as well. 

 

To generate the onscreen text response, simply hold down the keybind you customized in the INI file and speak your desired question from either of the Responses xml files, or any questions of your own that you created, and then let off the keybind for the text response to be displayed onscreen. Be sure you are holding down the keybind prior to speaking the phrase and not letting off until the entire phrase has been spoken, so that the proper response is triggered. Don't be afraid to experiment with with the phrases. If you have trouble getting the phrases to be recognized, change them up. Find the ones giving you trouble and customize them with a different phrase. It will require some trial and error, but once you get it sorted, you interactions will be much more consistent. If you're having speech recognition issues, adjust the new confidence threshold option in the INI. If that's still not working for you, make sure you try both engine options (WSR & MSP). If it's still not working, then it's an issue on your end, not with NPCI. NPCI does not control speech recognition, Windows does. NPCI only enables your audio to pass through, use your local speech recognition to match your spoken phrase, and then display the onscreen text it matches with from the XML's. 

 

Remember, many of these questions/phrases you already RP in-game each time you interact with a PED. Just modify the ones I provided to match the way you like to speak those commonly used phrases! It will help to cut down on memorization. 

 

Don't overlook the option to customize the unrecognized text responses that show onscreen for unrecognized input. Also, the unrecognized responses now have PED animations as well, to make the unrecognized responses feel more like a part of the interaction and less like an error. If you customize these just right, these unrecognized responses can fit a lot of scenarios and feel more like deliberate responses, making it easier for you to continue with your PED interactions! Come up with phrases/responses to ask back to the PED when getting an unrecognized response. This will help play out those instances and make the responses feel like they're part of the interaction. 

 

Suggestions & Bug Reporting:

All bug reports and suggestions will be done through Discord. Join my Discord server here: https://discord.gg/9RMrrvm5qH

 

Future releases

• Add audio voices for the responses (with optional onscreen text)

 

• Add in more PED and Player action animations

 

• Future enhancements using AI interaction for more natural, conversational, and unscripted exchanges. 

 

© 2025 NPCI by doublehook. All rights reserved. Do NOT reverse engineer, modify, or share the contents of my plugin without my express written consent

 

 

 

Edited by doublehook

What's New in Version 2.1.0.0

Released

* DUAL speech recognition support. Set the engine to use WSR (legacy Windows speech recognition) or MSP (the new Microsoft Speech Platform) in the NPCI.INI file

* New speech recognition confidence threshold customization - modify the confidence threshold (WSR or MSP) to better match your speech recognition needs

* New blacklist option for all emergency personnel (default is enabled)

* New PED actions by voice (PED wait by patrol vehicle & PED go to crosshair location & more)

* New Player actions by voice

* New PED automark feature for both PR & STP - marks the PED automatically for you when giving the frisk action command (in case you forget prior to giving the command)

* XML update that brings nested phrases into a single command - to help better organize your XML files and alleviate duplicate responses for each unique phrase

* Controller support for NPCI voice input and for the new PedGoToCrosshair action (customizable in the NPCI.INI file)

 

 

 


Short Description

A brand new way to talk to PEDS in LSPDFR - using just your voice!

User Feedback

Recommended Comments

doublehook

Members Author

12 minutes ago, Erekle said:

Could contain: text, clothing, person, screenshot, dress, poster, pattern (fashion design)

Hey Erekle - I just sent you a follow up email. As soon as I get your rage log, I should be able to pinpoint the issue. Thanks.

Erekle

Members

10 hours ago, doublehook said:

Hey Erekle - I just sent you a follow up email. As soon as I get your rage log, I should be able to pinpoint the issue. Thanks.

I sent an email about the rage log

SnakeDoc29

Members

So what script do you use to make your players mouth move?

zephynx

Members

it doesn't recognize anything i say, just sends a generic response, and after a while, my whole gta crashes as a result of it, any tips?

doublehook

Members Author

3 hours ago, zephynx said:

it doesn't recognize anything i say, just sends a generic response, and after a while, my whole gta crashes as a result of it, any tips?

Hi zephynx. I'm sorry for the issues you're currently experiencing. I notice your profile location says Malaysia, which doesn't officially support the built-in Windows Speech Recognition (which NPCI uses). Are you running an English version of Windows? If you are running Windows en-US, can you please fill out my Google form here (https://forms.gle/vrgGLogwacrFMqyw8) and provide your email address, so I can work with you through my support portal. Thanks in advance. 

1 minute ago, doublehook said:

Hi zephynx. I'm sorry for the issues you're currently experiencing. I notice your profile location says Malaysia, which doesn't officially support the built-in Windows Speech Recognition (which NPCI uses). Are you running an English version of Windows? If you are running Windows en-US, can you please fill out my Google form here (https://forms.gle/vrgGLogwacrFMqyw8) and provide your email address, so I can work with you through my support portal. Thanks in advance. 

Just FYI - I will be porting NPCI over to the new "Microsoft Speech Platform" soon, which is Windows new speech recognition platform, and does support Malay. 

zephynx

Members

12 hours ago, doublehook said:

Hi zephynx. I'm sorry for the issues you're currently experiencing. I notice your profile location says Malaysia, which doesn't officially support the built-in Windows Speech Recognition (which NPCI uses). Are you running an English version of Windows? If you are running Windows en-US, can you please fill out my Google form here (https://forms.gle/vrgGLogwacrFMqyw8) and provide your email address, so I can work with you through my support portal. Thanks in advance. 

Just FYI - I will be porting NPCI over to the new "Microsoft Speech Platform" soon, which is Windows new speech recognition platform, and does support Malay. 

Yea, I'm speaking English for this plugin by the way. Not sure if this matters, I'll fill up the form now.

As i just found this plugin And i recently Found out that MSR isn't support and windows forced me onto Version 24H2 Is this plugin Unusable now?

 

2nd question Are you going to be adding Support for policing Redefined  When it release?

 

doublehook

Members Author

(edited)

8 hours ago, LeArmouredbadger said:

As i just found this plugin And i recently Found out that MSR isn't support and windows forced me onto Version 24H2 Is this plugin Unusable now?

 

2nd question Are you going to be adding Support for policing Redefined  When it release?

 

Hi there. Thanks for your message. For your first question, WSR is still part of Windows 24h2, it's just not configurable like previous Windows versions. I just built a new computer a month ago and installed 24h2 right out of the gate, and I've had no issues using NPCI under WSR. My latest 2.0 demonstration video is using 24h2. I will eventually switch over from WSR (Windows Speech Recognition) to the new MSP (Microsoft Speech Platform), but for now, NPCI works just fine on the latest version of Windows. 

 

Regarding your 2nd question, the short answer is yes. My goal will be to make sure they work interchangeably with each other without any compatibility issues upon PR's release. I'm even delaying the release of NPCI 2.0 until I've had some time to test them together. 

 

NPCI's goal from the start has always been about avoiding menu's (when possible) and adding more immersion to your PED interactions. NPCI won't control any of Policing Redefined's upcoming features (that's their proprietary code), but it should be a great companion plugin to use alongside PR. Perhaps the PR authors might be interested in incorporating NPCI's voice capabilities into their plugin down the road. Who knows. Could make for a powerhouse plugin!  🙂

 

I'm very excited for the upcoming release of Policing Redefined! I too want to be able to use both PR and NPCI together! I hope this answers your questions.  

Edited by doublehook

On 9/17/2025 at 1:50 AM, doublehook said:

Hi there. Thanks for your message. For your first question, WSR is still part of Windows 24h2, it's just not configurable like previous Windows versions. I just built a new computer a month ago and installed 24h2 right out of the gate, and I've had no issues using NPCI under WSR. My latest 2.0 demonstration video is using 24h2. I will eventually switch over from WSR (Windows Speech Recognition) to the new MSP (Microsoft Speech Platform), but for now, NPCI works just fine on the latest version of Windows. 

 

Regarding your 2nd question, the short answer is yes. My goal will be to make sure they work interchangeably with each other without any compatibility issues upon PR's release. I'm even delaying the release of NPCI 2.0 until I've had some time to test them together. 

 

NPCI's goal from the start has always been about avoiding menu's (when possible) and adding more immersion to your PED interactions. NPCI won't control any of Policing Redefined's upcoming features (that's their proprietary code), but it should be a great companion plugin to use alongside PR. Perhaps the PR authors might be interested in incorporating NPCI's voice capabilities into their plugin down the road. Who knows. Could make for a powerhouse plugin!  🙂

 

I'm very excited for the upcoming release of Policing Redefined! I too want to be able to use both PR and NPCI together! I hope this answers your questions.  

Hey thanks for Answering my Questions i really do hope you and PR could Work together to Combine both mods because that would be really Great and something that is desperately needed

crawfishdakilla

Members

This doesn't seem to be recognizing my speech, as I get a generic response no matter what. Did I miss something that I shouldve downloaded? My grammar police works fine and can hear me but for some reason this can not 

doublehook

Members Author

(edited)

2 hours ago, crawfishdakilla said:

This doesn't seem to be recognizing my speech, as I get a generic response no matter what. Did I miss something that I shouldve downloaded? My grammar police works fine and can hear me but for some reason this can not 

Hey there. If you don't mind, fill out my Google form here (https://forms.gle/vrgGLogwacrFMqyw8), so I can get your email address and hit you up for a copy of your rage log to see what errors it's generating. To answer your question, you don't need to install anything additional. It uses WSR (Windows Speech Recognition), but I recently installed a fresh copy of Windows 24h2 on a new PC and had no problems getting WSR to recognize my speech input even though it's being phased out for MSP (Microsoft Speech Platform). Since your profile shows you as USA, most likely using en-US language, you shouldn't have any issues. 

Edited by doublehook

AngelRelck

Members

Can i use Policing Redefined with this mod or i need STP? 

doublehook

Members Author

(edited)

10 hours ago, AngelRelck said:

Can i use Policing Redefined with this mod or i need STP? 

Hey there. I'm going to release 2 versions, one compatible w/STP, and one compatible w/Policing Redefined. I've spent the past several months working on v2.0 before PR came out, and there are several differences between STP & PR that will require different builds. To my surprise, NPCI v2.0 works much better with PR, as their code is much more refined and easier to play along with then the very outdated STP. It works so well in fact, that I had to make several code changes, as things not fully functioning with STP are now possible with PR. Bottom line, you'll have a version to use with whichever one you decide to go with.  😀 

Edited by doublehook

Awall

Members

Hi! This is an amazing plugin, how is the 2.0 version coming along? Do you have an approximate release date for a beta?

doublehook

Members Author

5 hours ago, Awall said:

Hi! This is an amazing plugin, how is the 2.0 version coming along? Do you have an approximate release date for a beta?

Hey there! Thanks so much for the kind words. Version 2.0 is basically done. I've been testing non-stop making little tweaks along the way. I'm currently finishing the user documentation and hope to release it in the next 7 to 10 days. I promise, it won't be much longer! Are you using Policing Redefined or STP? 

impliedstone

Insiders

such a good idea ..just does not work well yet, i just always get the default response ... takes away from the immersion when you have to repeat yourself 4 times ...i really hope this ends up good because it is sooooooo needed....

doublehook

Members Author

(edited)

50 minutes ago, impliedstone said:

such a good idea ..just does not work well yet, i just always get the default response ... takes away from the immersion when you have to repeat yourself 4 times ...i really hope this ends up good because it is sooooooo needed....

Thanks for your comment. I'm sorry it's not working out for you as good as you hoped. I've been using it for almost 8 months, and I can tell you there is a learning curve, and it takes a little time and practice to get the feel. After a little time though, you'll find it works quite well. Just make sure you're holding down your key bind before you speak, not during or at the same time, and don't let off until the full phrase is spoken. Once you get acclimated to the phrases, your experiences will be much better. I think that's the hardest part of the NPCI learning curve - memorizing the phrases. Also, don't be afraid to customize the "trigger" phrases to something that flows better for you. You can fully customize everything inside of the XML files to suit your style of interaction. I hope your experiences improve. Don't give up! One last note -- the F4 console shows your confidence level when you speak a phrase (for the ones it registers). If it's not above say 75%, you might try making some adjustments to your mic settings to see if that helps. 

Edited by doublehook

1 hour ago, doublehook said:

Thanks for your comment. I'm sorry it's not working out for you as good as you hoped. I've been using it for almost 8 months, and I can tell you there is a learning curve, and it takes a little time and practice to get the feel. After a little time though, you'll find it works quite well. Just make sure you're holding down your key bind before you speak, not during or at the same time, and don't let off until the full phrase is spoken. Once you get acclimated to the phrases, your experiences will be much better. I think that's the hardest part of the NPCI learning curve - memorizing the phrases. Also, don't be afraid to customize the "trigger" phrases to something that flows better for you. You can fully customize everything inside of the XML files to suit your style of interaction. I hope your experiences improve. Don't give up! One last note -- the F4 console shows your confidence level when you speak a phrase (for the ones it registers). If it's not above say 75%, you might try making some adjustments to your mic settings to see if that helps. 

I’m updating now didn’t realize npci was parsing these results !! You are a king bro 

YUGEKUNAMI

Members

Wow, the preview alone is amazing! great work and development, can't wait to try this! Thank you for releasing this for the community. 

doublehook

Members Author

26 minutes ago, YUGEKUNAMI said:

Wow, the preview alone is amazing! great work and development, can't wait to try this! Thank you for releasing this for the community. 

Thank you so much for the kind words. The LSPDFR community has truly enriched my life over the past year, and I'm so grateful I'm able to give something back to show my gratitude. Hopefully my contributions will have the same impact on someone else.

 

The beta testing is going very well, so the launch day will be soon!!

nova2727

Members

What be great is to have the peds hug you if it’s a person in distress and you calmed the person down or shake hands with you like a witness and after you take information they thank you and shake u hand

doublehook

Members Author

(edited)

2 hours ago, nova2727 said:

What be great is to have the peds hug you if it’s a person in distress and you calmed the person down or shake hands with you like a witness and after you take information they thank you and shake u hand

I actually have a "bro" hand shake and hug action in version 2.0. They're not as formal as the ones you're describing, and sadly I don't think any exist for GTA 5, but for the ones that I've included in 2.0, you could script your interaction to achieve this (you'd modify the XML phrases and the desired responses). The PED won't do it automatically when you walk up to them though, as my plugin relies on voice input to trigger an action.

 

I'll double check to see if there are more formal hand shakes and hug actions available in the GTA 5 animations library. If I can find some others that might better match your scenario, I'll add them in before the release. I'll put up a video showing the current actions, so you can at least see what I have in the plugin currently. I'll put it up on YT in the next day or so. Thanks for your suggested feature! 

Edited by doublehook

lspdfr swat fan

Members

Does this work with Policing Redefined's felony stop function? If not, could you add that as a feature?

Woensel

Members

The talking works perfectly fine but if i ask to do something or give their id or license the npc responds but doesnt do it. What can that be?

 

Create an account or sign in to comment

Latest Mods

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.