Repository files navigation

ChatGPT Android demonstratesOpenAI's ChatGPT on Android withStream Chat SDK for Compose.

The primary objective of this repository is to showcase the following:

• DemonstratingOpenAI's chat APIs.
• Creating comprehensive UI elements utilizing Jetpack Compose.
• Incorporating Android architecture components using Jetpack libraries such as Hilt and AppStartup.
• Executing background tasks efficiently through Kotlin Coroutines and WorkManager.
• Seamlessly integrating chat systems via theStream Chat SDK for Compose to facilitate real-time event handling.

ChatGPT Android is built withStream Chat SDK for Compose to implement messaging systems. If you’re interested in adding powerful In-App Messaging or real-time Video Chat to your app, check out theAndroid Chat Messaging Tutorial orAndroid Video Call Tutorial !

• Stream Chat SDK for Android on GitHub
• Android Samples for Stream Chat SDK on GitHub
• Stream Chat Compose UI Componenets Guidelines

If you're interested in building a real-time meeting room with video/audio calling, audio room, or livestreaming, check out the references below:

• Stream Video SDK for Android on GitHub
• Video Call Tutorial /Audio Room Tutorial /Livestream Tutorial

For proper project setup, it's essential to follow the instructions below. In most cases, you can complete all configuration steps within 10 to 20 minutes:

1. Go to theStream login page.

2. If you have your GitHub or Google account, click theContinue with GitHub orContinue with Google button and you can sign up within a couple of seconds.

3. Go to theDashboard and click theCreate App button like the below.

4. Fill in the blanks like the below and click theCreate App button.

5. You will see theKey like the figure below and then copy it.

6. Create a new file namedsecrets.properties on the root directory of this Android project, and add the key to thesecrets.properties file like the below:

STREAM_API_KEY=..

7. Go to yourDashboard again and click your App.

8. In theOverview menu, you can find theAuthentication category by scrolling to the middle of the page.

9. Switch on theDisable Auth Checks option and click theSubmit button like the figure below.

11. Click theExplorer tab on the left side menu.

12. Clickusers ->Create New User button sequentially and add fill in the user like the below:

• User Name:Chat GPT
• User ID:70ef052a-da88-4451-af92-99f7ed335a71

13. Now, it's time to set up the OpenAI API key. Go toOpenAI API Dashboard, login with your Google account and clickStart verification to verify your phone number to create an API key.

14. After verifying your phone number, click theCreate new secret key on the page like the image below and create your API key:

15. Copy and paste your API key to thesecrets.properties file as namedGPT_API_KEY under theSTREAM_API_KEY field:

GPT_API_KEY=..

16. Unlock Rate limits for OpenAI APIs

As per the guidelines outlined in theirRate limits documentation, access to the free tier of the API isgranted after a minimum expenditure of $5 since the creation of your account. That means, you should register your payment method to the OpenAI and purchase initial credits.

So you should go to theOpenAI Billings and click theAdd payment details button like the image below:

17. Once you've added your payment deatils, you'll able to see theAdd to creadit balance button like the image below:

Next, click theAdd to credit balance button to purchase the initial credit. As you've discovered in the previous step, the minimum credit starts from $5. Therefore, if you only intend to utilize the API limits for testing or building a side project, you can opt for the minimum pricing. Also, not that it's advisable to disable the automatic recharge option to avoid unexpected charges.

18. Once you've completed all those steps above, now build, run and enjoy your the project!

You can typically find the reasons behind your error code by referring to theOpenAI Error code documentation.

If you encounter 401 errors while chatting with ChatGPT app, as shown in the image below, it's crucial to verify the correctness of your OpenAI API key. Ensure that there are no typos in your API key and that it has been correctly copied and pasted into thesecrets.properties file, as outlined in the guidelines above.

If you encounter this error from the outset, it's likely that your rate limits do not meet the requirements. Please revisit step 16 in the guidelines above for further instructions.

• Minimum SDK level 21.
• 100%Jetpack Compose based +Coroutines +Flow for asynchronous and cold streams.
• Compose Chat SDK for Messaging: The Jetpack Compose Chat Messaging SDK is built on a low-level chat client and provides modular, customizable Compose UI components that you can easily drop into your app.
• OpenAI Chat API: Given a list of messages comprising a conversation, the model will return a response.
• Jetpack
  • Compose: Android’s modern toolkit for building native UI.
  • ViewModel: UI related data holder and lifecycle aware.
  • App Startup: Provides a straightforward, performant way to initialize components at application startup.
  • Navigation: For navigating screens andHilt Navigation Compose for injecting dependencies.
  • Room: Constructs Database by providing an abstraction layer over SQLite to allow fluent database access.
  • Hilt: Dependency Injection.
  • WorkManager: To achieve background processing with scheduling.
• Landscapist Glide,animation,placeholder: Jetpack Compose image loading library that fetches and displays network images with Glide, Coil, and Fresco.
• Retrofit2 & OkHttp3: Construct the REST APIs and paging network data.
• Sandwich: Construct a lightweight and modern response interface to handle network payload for Android.
• Moshi: A modern JSON library for Kotlin and Java.
• ksp: Kotlin Symbol Processing API.
• Balloon: Modernized and sophisticated tooltips, fully customizable with an arrow and animations for Android.
• viewmodel-lifecycle: ViewModel Lifecycle allows you to track and observe Jetpack's ViewModel lifecycle changes.
• StreamLog: A lightweight and extensible logger library for Kotlin and Android.
• Baseline Profiles: To improve app performance by including a list of classes and methods specifications in your APK that can be used by Android Runtime.

ChatGPT Android follows theGoogle's official architecture guidance.

ChatGPT Android was built withGuide to app architecture, so it would be a great sample to show how the architecture works in real-world projects.

The overall architecture is composed of two layers; UI Layer and the data layer. Each layer has dedicated components and they each have different responsibilities.The arrow means the component has a dependency on the target component following its direction.

Each layer has different responsibilities below. Basically, they followunidirectional event/data flow.

The UI Layer consists of UI elements like buttons, menus, tabs that could interact with users andViewModel that holds app states and restores data when configuration changes.

The data Layer consists of repositories, which include business logic, such as querying data from the local database and requesting remote data from the network. It is implemented as an offline-first source of business logic and follows thesingle source of truth principle.

For more information about the overall architecture, check outBuild a Real-Time WhatsApp Clone With Jetpack Compose.

ChatGPT Android adopted modularization strategies below:

• Reusability: Modulizing reusable codes properly enable opportunities for code sharing and limits code accessibility in other modules at the same time.

• Parallel Building: Each module can be run in parallel and it reduces the build time.

• Decentralized focusing: Each developer team can assign their dedicated module and they can focus on their own modules.

Most of the features are not completed except the chat feature, so anyone can contribute and improve this project following theContributing Guideline.

Support it by joiningstargazers for this repository. ⭐
Also,follow me on GitHub for my next creations! 🤩

Designed and developed by 2022 skydoves (Jaewoong Eum)Licensed under the Apache License, Version 2.0 (the "License");you may not use this file except in compliance with the License.You may obtain a copy of the License at   http://www.apache.org/licenses/LICENSE-2.0Unless required by applicable law or agreed to in writing, softwaredistributed under the License is distributed on an "AS IS" BASIS,WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.See the License for the specific language governing permissions andlimitations under the License.