When Google denote Gemini , I was frantic to attempt it because it is a multi - average good example that works with schoolbook , image , code , audio , and picture . Unlike ChatGPT , it can also get at the net and provide real - fourth dimension selective information . However , after using it for over six month , I am moving back to ChatGPT .
On paper , the late Gemini 1.5 Pro model outperforms OpenAI ’s young ChatGPT 4.0 . First , it supports double the keepsake for the maximal succession length compared to ChatGPT , which means it can understand the context of use of prompts much best . It also supports up to a 2 million token context windowpane , allowing it to keep previous chats and upload file cabinet in memory while giving a reaction .
Check this table for reference book :
mark that these details are revealed by the various company . The actual experience of using them would be different for most people . Here ’s my experience in realness :
mesa of Contents
1. Gemini App Replaces Google Assistant on Android
When you instal the Gemini app on Android , you are thrust to expend Gemini as your digital assistant , replacing the familiar Google Assistant . This variety is n’t welcome , as Gemini lack many features you utilize daily with Google Assistant .
For starters , I can not control my smart home , custom commands to control apps no longer put to work , and even recess features like the ability to describe Sung dynasty by humming are missing . to boot , Gemini struggles to understand what I am saying and operates much slower compared to Google Assistant . That ’s possible since Gemini is still “ training itself ” using my voice bid . Google lately add up the ability to pose reminders via Gemini , this improvement was long delinquent .
To go on using Google Assistant , I have to forgo add the Gemini app and rely on its web variant . or else of replacing Google Assistant , why ca n’t Google integrate Gemini ’s ability into Google Assistant , likeApple enhanced Siri?For now , I reason out that Gemini does not have a proper app on Android and has many downside .
If Gemini replaces Google Assistant on your phone , you canswitch back to Google Assistant from Gemini .
2. Gets the Math Wrong Most of the Time
Recently , I moved to a new way with three friends . We pay for expenses for the house and wanted to split them evenly at the end . I gave Gemini details about the brokerage house , progression , tape transport , and who paid how much , need it to look who should devote whom and how much to part the expense equally .
Here ’s the command prompt I have given to Gemini .
Despite bring home the bacon all the details to Gemini , it whole messed up the calculations . It suggested that my acquaintance and I had to pay the same amount to the third mortal , even though my acquaintance had paid more ab initio . to boot , it stated that Shyam would meet ₹ 18,900 , but did n’t explicate who should pay him . Even after several back - and - forth messages to correct its calculation , I could n’t get an exact answer .
I finally had to revert to ChatGPT to get the job sorted . ChatGPT provided a sporting and correct computing with the same prompt . Not only was the resolution accurate , but the explanation was also elaborated and clear .
This is n’t just a one - clock time issue . We alsotried ask math dubiousness , like solving arithmetic cognitive operation and observe derivative of functions , where Gemini failed most multiplication when the questions got complicated . It ’s not just math ; even with riddles or anything postulate logical thinking and logic , Gemini has a higher chance of failure . It sense like Gemini is n’t quick for the everyday maths yet .
3. Inconsistent and Incomplete Code Generation
I do n’t program daily , but I now and then wish to trifle with genus Apis or learn what ’s raw as an enthusiast . With my limited testing , I can say Gemini has grown much well than its initial Bard days . However , ChatGPT still outshine Gemini in proficiency .
For example , I gave a elaborate prompting to both Gemini and ChatGPT to create an disbursement tracker app using HTML , CSS , and JavaScript :
Both models skipped the editing expenses part , but Gemini did n’t even remark it . Additionally , to set it up and get the output , I had to manually specify many underage wrongdoing with Gemini ’s result .
The major issue for me with Gemini is that it lacks context when generating multiple codes . For example , I expect it to generate a login page , which it did successfully . However , when I asked it to generate the server - side script to handle the login logic , it created code with Modern variable that did not jibe the balance of the code . While the yield code was technically correct , it left more oeuvre and potential error for me to resolve , which was n’t the case with ChatGPT for the most part .
4. Lack of Explanation and Proper Structure
While Gemini from time to time provide a detailed explanation , it often skip this all-important part and startle flat to the resolution . Whether I ask it to generate code , solve a maths problem or riddle , or explicate a matter , I frequently find oneself myself needing to request further clarification because the initial answer is n’t clear or well - structured .
For example , this is the prompt I have given :
Here ’s the response from Gemini :
And this is what I mystify from ChatGPT :
Even when Gemini provides a elaborate step - by - footfall explanation , I ’ve found that the structure of its results is n’t as polished as ChatGPT ’s . In my experience , using ChatGPT often leads to a more cohesive and clearer introduction , although this druthers may motley from person to mortal .
5. Image Recognition Isn’t on Par with ChatGPT
In my usage of both avail , I ’ve uploaded various types of figure for different purposes — memes for explanation , screenshots of error content for issuance resolution , and photos of objects to garner details . From my experience , ChatGPT excels in understanding images and retrieving relevant information . Google does not leverage it ’s Google Lens capableness which often does a better line of work than Gemini itself when it come to find details about any target , animal , or plant .
Another significant issue with Gemini is its inability to get at image or Indian file uploaded in former content . For representative , I put up Gemini with a list of math problems in an image and postulate for the solvent to the first one , which it accurately put up . However , when I bespeak the solution for the second problem , Gemini seemed unaware of what I was referring to . Even when cue to refer back to the figure of speech sent in the previous content , Gemini was unable to access it . I test this with multiple ikon and file , and Gemini consistently could only study image uploaded within the same message .
what is more , there are instances where Gemini does n’t acknowledge uploaded images at all . For lesson , when I upload an image of a smartphone and asked for its model identification number , Gemini react with inside information about its own Gemini model rendering rather . This return stay across both the free and advanced variation of Gemini .
6. Other Minor Issues With Gemini
7. Missing Features on Gemini
Gemini lacks several feature of speech that are usable on ChatGPT :
Each of these features can be a major cause to choose ChatGPT over Gemini , as they conjointly offer a more comprehensive , drug user - favorable , and flexible experience .
Moving Forward
After my campaign - Hoosier State with Gemini for a prospicient time , I will rejoin to using ChatGPT as my primary quill go - to chatbot . I prefer using Perplexity orother Perplexity alternativesfor research because it has material - time info and explains in wide-eyed bulleted points . However , I will still be using Gemini for prize tasks and aim .