ReOldAI

Unofficial Gemini client for Android 2.3+

"Gemini" and other related trademarks are trademarks of Google. Google is not associated with ReOldAi or YMP Yuri.

Website • Boosty • YouTube: YMP Yuri

⬇️ DOWNLOAD

Download on Github Releases

Table of Contents

1. About the Project

2. Technologies

3. Installation

   • Requirements

   • Download

4. Usage

5. Screenshots

6. Compatibility Table

7. Table Explanation

8. Voice Input

9. How to Get a Gemini API Key

10. build.gradle Configuration

11. Roadmap

12. Contacts

13. License

14. Other Information

---

About the Project

ReOldAi is a lightweight and minimalistic application that allows interaction with Google AI's Gemini model on older Android devices (starting from version 2.2*, API 8).

Its minimal interface and lack of unnecessary dependencies allow it to run even on weak and outdated smartphones or tablets.

* - Be sure to check the compatibility table!

----------

Brief App Information

• Language: Java

• UI: Android Support Library (AppCompat v7, Support v4)

• Uses Conscrypt library (SSL/TLS) for new certificate support

• JSON Serialization: Google GSON

• Icons: Material Icons (Google)

• Languages: Russian and English

• Neural Network Used: Gemini API (Google AI Studio)

---

Installation

Requirements

• Device with Android 2.2 or higher (minimum API 8 support)

• Stable internet connection for Gemini API operation

• Gemini API key (see section below)

Download

• Link to Github releases: https://github.com/YMP-CO/ReOldAi/releases

---

Usage

1. Launch ReOldAi

2. Enter your Gemini API key in the window or settings

3. Start a dialogue in the text chat It's simple!

---

Screenshots

---

Compatibility Table

Feature	Android 2.1<	Android 2.2	Android 2.3	Android 3.0–3.2	Android 4.0	Android 4.1	Android 4.2>
Result	🟥 Even basic libraries are not compatible	🟧 App launches, but Conscrypt doesn't work due to library incompatibility	🟩	🟩	🟩	🟩	🟩
App Launch	🟥	🟩	🟩	🟩	🟩	🟩	🟩
Stability	🟥	🟨 (Libraries compatible only with Android 2.3)	🟩	🟩	🟩	🟩	🟩
Neural Network Operation	🟥	🟥 (Certificate error)	🟩	🟩	🟩	🟩	🟩
Voice Input	🟥	🟥	🟥	🟥	🟧 (Offline Google speech recognition support can be enabled)	N/A	🟩 (Requires Google app (Velvet.apk))
Message Copying	🟥	🟩	🟩	🟩	🟩	🟩	🟩

• 🟩 Full support
• 🟧 Partial support (works, but with limitations)
• 🟨 Potential crashes (limited stability)
• 🟥 Not supported
• N/A Not tested

---

Table Explanation

• Result: overall information on how the application functions.

• App Launch: whether the application starts correctly.

• Stability: number of critical crashes and freezes.

• Neural Network Operation: successful requests and receiving responses from Gemini.

• Voice Input: usage of Android Recognizer Intent (more details below).

• Message Copying: availability of the context menu and clipboard.

---

Voice Input

private void startVoiceInput() {
    Intent intent = new Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH);
    intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE_MODEL, RecognizerIntent.LANGUAGE_MODEL_FREE_FORM);
    intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE, Locale.getDefault());
    intent.putExtra(RecognizerIntent.EXTRA_PROMPT, "Say something...");

    try {
        startActivityForResult(intent, SPEECH_INPUT_REQUEST_CODE);
    } catch (ActivityNotFoundException e) {
        Toast.makeText(getApplicationContext(),
                "Your device does not support voice input.",
                Toast.LENGTH_SHORT).show();
    }
}

This feature requires Google app version 3.0 or higher (not verified)

---

How to Get a Gemini API Key

1. Go to Google AI Studio and sign in with your Google account.

2. Click the "Get API KEY" button at the top.

3. Then, click "Create API key".

4. Done!

---

build.gradle Configuration

plugins {
    alias(libs.plugins.android.application)
}

android {
    namespace 'com.ymp.reold.ai'
    compileSdk 34

    defaultConfig {
        applicationId "com.ymp.reold.ai"
        minSdk 8
        targetSdk 34
        versionCode 1
        versionName "1.0"

        testInstrumentationRunner "androidx.test.runner.AndroidJUnitRunner"
    }

    buildTypes {
        release {
            minifyEnabled false
            proguardFiles getDefaultProguardFile('proguard-android-optimize.txt'), 'proguard-rules.pro'
        }
    }
    compileOptions {
        sourceCompatibility JavaVersion.VERSION_11
        targetCompatibility JavaVersion.VERSION_11
    }
}

dependencies {
    implementation 'com.android.support:appcompat-v7:23.0.0'
    implementation 'com.android.support:support-v4:23.0.0'

    implementation 'org.conscrypt:conscrypt-android:2.5.2'
    implementation 'com.google.code.gson:gson:2.8.6'

    testImplementation 'junit:junit:4.12'
    androidTestImplementation 'com.android.support.test:runner:0.5'
    androidTestImplementation 'com.android.support.test.espresso:espresso-core:2.2.2'
}

---

Roadmap

• Finish Russian interface translation

• Add support for additional Gemini models

• Expand neural network parameter settings (top-p, temperature)

• Implement dark theme

• Oh, how about I add support for Android 1.0+ (in my dreams)

---

Contacts

• YouTube: YMP Yuri

• Boosty: gachic_the_geek

• Website: ymp-co.github.io

---

License

Distributed under the CC-BY 4.0 License.

---

Other Information

Two methods were used to enable message copying:

private void copyTextToClipboard(String text) {  
    // For Android API 11 (Honeycomb) and higher  
  if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.HONEYCOMB) {  
        ClipboardManager clipboard = (ClipboardManager) getContext().getSystemService(Context.CLIPBOARD_SERVICE);  
        ClipData clip = ClipData.newPlainText("AI Chat Message", text);  
        clipboard.setPrimaryClip(clip);  
    }  
    // For Android API 8, 9, 10 (Froyo, Gingerbread ->)  
  else {  
        //android.text.ClipboardManager  
  android.text.ClipboardManager clipboard =  
                (android.text.ClipboardManager) getContext().getSystemService(Context.CLIPBOARD_SERVICE);  
        clipboard.setText(text);  
    }  
    Toast.makeText(getContext(), "Copied text.", Toast.LENGTH_SHORT).show();  
}

Also, I want to warn everyone that I am only 16 years old. I am not a professional programmer; I used neural networks for development, etc. Please don't judge too harshly - the project is in beta testing!