Exploring NVDA for Android: Leveraging Open-Source TalkBack Architecture and Native Engine Integration

[ghost]

Hi NV Access Community & Developers,

I would like to open a technical discussion regarding the feasibility and community interest in bringing NVDA concepts or an official NVDA ecosystem to Android.

While I understand that NVDA is primarily built for Windows using Python, expanding accessibility options on mobile is becoming increasingly critical. Since Android's TalkBack code is open-source and hosted on GitHub (available at https://github.com/google/talkback), its architecture can serve as a strong foundation or reference point for building a dedicated accessibility service.

Beyond expanding accessibility, bringing NVDA to Android offers powerful unique benefits for users:

1. Seamless Settings Synchronization: Users could sync their speech rates, voice profiles, dictionaries, and key bindings between Windows and Android.
2. Unified User Experience: Visually impaired users would enjoy a familiar keyboard navigation structure and shortcut philosophy whether using a PC or a Bluetooth keyboard with their Android phone.
3. Open-Source Engine Integration: Native speech engines like eSpeak NG can be compiled and integrated seamlessly with Android's TextToSpeechService API to maintain fast response times and low latency. I have already personally worked on building and compiling eSpeak NG for Android with success.

Having an official community initiative or guidance from NV Access to bridge these open-source tools on Android would be a huge leap for visually impaired users worldwide.

I would love to hear the community's and developers' thoughts on how we can collaborate on this.

Best regards,
uyt3

[seanbudd]

NV Access is not looking to support Android. Writing an android screenreader would have to be done from scratch, you cannot port NVDA to a different OS. Screen readers are extremely OS specific. I would encourage starting a new separate project or fork for android unrelated to NVDA and NV access

[rk974110]

if open source screen reader like NVDA came on android, it can beat so many screen readers who relys on talkback source code and customises on there own and spreading in the community.

[ghost]

@rk974110 You are absolutely right. This will also unify the experience between phones and computers. Mobile accessibility is no less important than computer accessibility. If you know anyone who can collaborate, share the link to this discussion with them so we can cooperate and help this organization establish this project.

[hemanthmuniraju]

It is a great idea to try to bring NVDA screen reader to Android. Wonderful thinking. If it comes true, really many screen reader users will be happy, and even they can have more control on their smartphone without any fear, without any hesitation, with the help of an open source method. Wonderful idea. Try to bring it as soon as possible.

[ghost]

@hemanthmuniraju Thank you for your strong support!
To achieve this vision, we need to spread this discussion as much as possible among developers and the accessibility community. Our primary goal is to demonstrate the huge demand to NV Access so they will establish an official, documented repository for an Android project. Official hosting by NV Access is key for credibility.
My core goals with this initiative are:

1. Cross-Platform Sync: Enabling seamless synchronization of settings and speech dictionaries between Windows and Android for a unified user experience.
2. Honoring Our History & Spurring Innovation: Remember the early days when open solutions like INF files and Windows Talking enabled visually impaired users to independently install Windows from scratch? That same spirit of total control and freedom is what we need on Android today. Just as Microsoft’s rapid development of Narrator created healthy competition for NVDA on PC, bringing a true community-driven, open-source engine to Android will break current limitations and push Google to significantly evolve TalkBack.
   Please share the link with any developers who might be interested so we can build momentum together!

[ghost]

Hello @luiscarlos2000,
I completely agree with your opinion. I previously contacted several developers who mentioned they would join this discussion, though I am not sure why they haven't participated yet.
Here are a few leads and developers you might want to reach out to:

• @ferda (Developer of a screen reader):
  You can contact him on Telegram via his username: @mustafadogan7678.
• Team developing a screen reader & assistant:
  I know a team building a screen reader from scratch that also serves as an assistant for the blind. Although it is in its early stages and has some issues, it is very promising. You can contact one of the developers via Telegram: @Milan_thapa92.
• Talk Forward Developer:
  I previously reached out to the developer of Talk Forward, who holds an NVDA certificate and is a certified expert. I suggested certifying the reader through that institution, but he declined, stating the project is still in its early stages. If you wish, you can get in touch with his team members who are participating in this initiative: @kaillewaille.
• HearSee Screen Reader Developer:
  I tried contacting the developer of HearSee Screen Reader on Telegram previously, but he didn't reply. You can try reaching out to him via his username: @Erray0507.
  These are the key developers I personally know and have contacted on Telegram. I strongly encourage everyone to share this discussion link so more people can join us and help make this initiative a success!

[DrAttia]

Extremely important, i think it is very promising for the accessibility community.

[ghost]

Yes, you are very right. You are right, it is a very important matter.
If we got more developers and we all agreed to form a team, it would be great.
@DrAttia

[wsl19951001]

Very good. I also support the idea

[ghost]

@wsl19951001 Yes, I also support this idea, which is why I started this discussion, hoping that someone might respond to us. Unfortunately, no one ever responds. I know it's not that easy, but I cannot do this alone. I need developers who want to do this, and I need this program to be implemented by this organization to sync settings and take advantage of the features they offer on the Windows system.
I will keep repeating it: mobile accessibility is not more important than computer accessibility.
I would like to have a unified experience for both mobile and desktop.
Greetings to everyone

[luiscarlos2000]

Hi,
Well, those are the devs for the talkback clones. From what I've researched, someone who I want... or don't want? to name (sorry sorry I'll do it!) is developing a screen reader for Android from scratch. One Akash Kakkar or something?
I've posted this discussion on the android forum for some feedback.

[ghost]

@luiscarlos2000 No, my dear, someone has developed a screen reader program built from scratch, and it is
Its developer's name is RippleLabs.
His name is Ray Bolt.
Sorry, his name is Rebolat.
As for the screen reader you are talking about, what is this screen reader? Tell me about it, I have not heard of it before.
I really wish they would create a screen reader for Android, my dear.

[luiscarlos2000]

Oh my, very confusing, it's very... very confusing! You're confusing that ray person with another, or maybe I'm missing something? Ok, about Akash's screen reader proposal, it's still in internal development, (he's using an internal codename). He will decide about a public name near release. Recently well few minutes ago I've asked him about it's progress since this discussion started.

[ghost]

@luiscarlos2000 I have not heard of this screen reader before; I would like to know who the developer is so that I can test it internally, if possible.
Also, tell me who the developer of this screen reader is, and whether it is open source or not.
I love constantly trying out new screen reader programs.
It seems the name was not clear enough. Is this screen reader open source or not?
And why, and why didn't the screen sharer participate here with us in this discussion?
Why didn't the screen reader developer join us in this discussion?

[luiscarlos2000]

Well, too many questions you have and too many answers I have, but since this discussion is going... beyond? my/our scope, maybe the BlindAndroidUsers forum on this topic (https://groups.io/g/blindandroidusers/topic/a_new_screen_reader_from/120500983) be the right way to discuss it Obviously account required, but with that you can have your ideas there.

Going to the answers to your questions. According on what I've heard, Do you remember the Audex app for Android? Well, that was it's first app, he is the same developer for his future screen reader. Right now it's only developed internally by him, but once a public test release is in, it will be open-source. Yes, that future android screen reader will be open source. As I've told you that screen reader doesn't have an official public name, only an internal codename that he uses and only he knows.
And about the developer... Sorry if I'm wrong, but maybe he doesn't have an account here on github so maybe he didn't joined here. But don't worry, maybe he's well known in other places even by me, so maybe we're still in contact from each other.

[ghost]

@luiscarlos2000 Okay, thank you. Let's not talk about this reader again, so that the discussion doesn't veer into the requested topic.
Thank you, and of course I will take a look at this topic on this forum. Thank you very much.

[josephsl]

Hi all,

Before we continue this discussion (here or somewhere else), let me pose four important questions that should be answered by anyone wishing to develop screen readers:

1. There was a discussion about borrowing the development philosophy of NvDA as part of an Android screen reader. Besides equal access to technology and accepting third-party contributions, are there ar ideas and philosophies from NVDA that are applicable?
2. Screen readers are complex software with many components working together to offer accessibility experience. Using NVDA's documentation and source code, describe at least four components of NVDA that are portable across operating systems and at least two components not portable across operating systems.
3. I understand that synchronizing screen reader settings and providing keyboard-centric accessibility across platforms are attractive ideas. Building from question 2, are there other compelling reasons for porting NVDA to Android? Hint: you CANNOT use screen reader usage statistics as one of your reasons (people have tried using numbers to argue for porting screen readers but we need something else).
4. As discussed by participants, there are other Android screen readers that could possibly incorporate features from NVDA in an Android-centric way. But why did you gravitate toward porting NVDA instead? Is it because of lack of features in other screen readers or development philosphy differences?

Something to think about: you asked why other screen reader developers are not participating here yet. Why do you think that is? Big hint: you have partly answered this question in your first post.

Please answer these questions before continuing.

Thanks.

[ghost]

@josephsl Let me tell you why we want to migrate this specific program: first, its response speed, and its many features that are not available in other screen readers, such as browse mode, among many others.
Regarding transportation, I know it is extremely difficult, but have you ever imagined the experience being unified on both mobile and PC at the same time?
This matter deserves more and more consideration.
Devices, as phones have now become more important than computers themselves.
This is from my personal point of view.
Since no one is interested, I have posted this discussion as an idea, and I am not forcing anyone at all to create this application.

[josephsl]

Hi,

I see the attractiveness of offering a unified experience on the mobile and PC side. However, I don't think you answered my questions to the fullest: I asked you to consider the components that make up NVDA and how these components can be ported or perhaps not portable.

The purpose of asking the technical questions was to see if you have thought about what it really takes to develop or port a complex software like a screen reader. If your reasons are strictly feature suggestions, then I think a more appropriate venue would be Android screen reader repositories and asking folks there to add such and such features. But since your main point is porting NVDA across operating systems (or at least collaborating with some folks to adopt NVDA's features in a cross-platform way), I thought you would have done some thinking about technical details. However, after reading the response, I think we all need to do some more work to truly understand how NVDA really works and design decisions made in the past.

One part of the response caught my attention: NVDA is responsive. While I agree with that statement from a user's point of view, this is not entirely the case once you figure out the actual structure of NVDA executable and its dependencies. Here is the truth: NVDA is a Python-based screen reader. I'll leave it up to you to figure out why this statement can question the assertion that NVDA is responsive (or at least, help you think a bit about what's going on underneath).

Another thing to think about: let's say NV Access and third-party contributors say "yes" to porting NVDA to Android. People tasked with porting NVDA must answer the following questions:

1. Is the Android NVDA going to use Python that can run on Android, or perhaps rely mostly on Java as its implementation language?
2. Which compiler chain would you use to compile Android NVDA or its equivalent? Remember that Microsoft Visual Studio (the toolchain NVDA people use) are primarily designed to compile binaries for the Windows operating system.
3. Which Android API are you going to target initially? Google changes Android API annually (or several times a year).
4. Can you keep up with constant updates for Android apps? One of the reasons why NVDA keeps updating its source code is to keep up with Windows app updates, including web browsers that can introduce subtle issues here and there. In particular, how will you keep up with mobile web browser updates should you choose to bring an Android implementation of NVDA's browse mode?
5. Which smartphones (and devices) are you going to prioritize when supporting Android NVDA for testing and real-life use? Unlike platforms like iOS, there are many manufacturers of Android devices and Android versions (and overlays in case of Samsung Galaxy series) varies. This also connects with a question above on which Android API version to target.
6. How will the Android NVDA keep up with its parent (Windows NVDA)? You need people to synchronize source code across ports. Ideally NVDA would support Windows and Android under a single source code base. But if people decide to create Android NVDA via forking, an additional obligation must be met: keep features synchronized across platforms. Unless the porting team decided to go their way and let Android NVDA diverge from Windows NVDA, the need to keep the source code up to date with the parent project becomes important.
7. Is keyboard the primary interaction modality? It is possible to run Android on a PC and use a keyboard to interact with Android smartphones and devices. However, we must not forget that the primary device form factor Android targets (even in 2026) is touchscreen equipped phones. Sure, Android NVDA can offer keyboard-centric Android accessibility. But the porting team must NOT forget the primary input modality on Android: touchscreens.

Even though our discussion is user-centric, as our discussion is taking place on a repository of a screen reader source code, I think some level of comfort with thinking in geeky ways could be helpful. This is why I am asking the above questions.

Hope this helps.

Thanks.

[seanbudd]

Locking. This thread appears to be brigaded by spam bots. We do not want unsupervised AI agents interacting with this repository. You should only comment as a human who has written or reviewed your message contents.