Skip to main content

Generative AI

Generative AI Revolution: A Comparative Analysis


In the world of Generative Artificial Intelligence (AI), a new era of large language models has emerged with the remarkable capabilities. ChatGPT, Gemini, Bard and Copilot have made an impact in the way we interact with mobile device and web technologies. We will perform a comparative analysis to highlight the capabilities of each tool.


Training DataWebWebWebWeb
F1 Score91%92%75%84%
InputsGPT-3.5: Text Only
GPT-4.0: Text and Images
Text, Images and Google DriveText and ImagesText and Images
Real Time DataGPT-3.5: No
GPT-4.0: Yes
Mobile SDK Only Only
Gemini Pro
Gemini Pro Vision

Calculation Metrics:

TP – True Positive

FP – False Positive

TN – True Negative

FN – False Negative

Accuracy = (TP +TN) / (TP + FP + TN + FN)

Recall = TP / (TP + FN)

Precision = TP / (TP + FP)

F1 Score = 2 * (Precision * Recall) / (Precision + Recall)


Our sample data set consists of 100 queries against Gemini AI. The above formula applied calculates the following scores:

Accuracy: (85 + 0) /100 = 85%

Recall: 85/ (85 + 5) = 94.44%

Precision: 85/ (85 + 10) = 89.47%

F1-Score: 2 * (0.894 * 0.944) / (0.894 + 0.944) = 91.8%


Recommended AI Tool:

I recommend Gemini based on its accuracy and consistency.  The ease of integration to the iOS and Android platforms and performance stands out amongst it’s competitors. We will illustrate how easy it is to integration Gemini with 10 easy steps.

Let’s Integrate Gemini into an Android Application!

  1. Download the Android Studio preview release Canary build (Jelly Fish| 2023.3.1).
  2. Create a new project: File -> New -> New Project
  3. Select the Phone and Tablet
    1. Under New Project -> Generate API Starter
    2. Click Next to Proceed
  4. Fill all the necessary details
    1. Enter the Project Name: My Application (or whatever you want to name your project)
    2. Enter the Package Name: (com.companyname.myapplication).
    3. Select the Location to save the project
    4. Select the Minimum SDK version: as API 26 (“Oreo”;Android 8.0)
    5. Select the Build Configuration Language as Kotlin DSL(build.gradle.kts)
    6. Click Finish to proceed 
  5. Create a starter app using the Gemini API
  6. To generate the API. Go to Gemini Studio.
  7. Click Get API Key -> click Create API Key in New Project or Create API Key in Existing Project in the Google AI studio
  8. Select the API key from the Prompt and paste in the Android Studio.
  9. Click Finish to proceed.
  10. Click the Run option in the Android Studio.

And you’re up and running with Generative AI in your Android app!

I typed in “Write a hello world code in java” and Gemini responded with code snippet. You can try out various queries to personalize your newly integrated Generative AI application to your needs.

Screenshot 2024 01 17 At 10.05.06 Pm

Alternatively, you can just Download the Sample app from the GitHub and add the API key to the to run the app.

It’s essential to recognize the remarkable capabilities of Generative AI tools on the market. Comparison of various AI metrics and architecture can give insight into performance, limitations and suitability for desired tasks. As the AI landscape continues to grow and evolve, we can anticipate even more groundbreaking innovations from AI tools. These innovations will disrupt and transform industries even further as time goes on.

For more information about Perficient’s Mobile Solutions expertise, subscribe to our blog or contact our Mobile Solutions team today!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Bhuvaneswari Kuduva Premkumar

Bhuvi is a Lead Technical Consultant on the Product Development team. She started her career as an iOS developer in 2015. As her journey evolved, she immersed herself in multifaceted cross-platform mobile application development, partnering with prominent clients in the retail and banking spheres.

More from this Author

Follow Us