Skip to main content

What LLM to use?

Introduction

It's been a few years since I started using large language models (LLMs) in the form of chat agents such as ChatGPT, Gemini, and DeepSeek. So far, they've been very helpful for me in so many areas. Mostly in building softwares, but they are too global in scope that their training are so helpful to so many people and so many subjects.

AI Chats

So far, I have these three in the list:

  1. ChatGPT
  2. Gemini
  3. DeepSeek
The longest I've used among these is ChatGPT. I've seen the best of its usefulness, and some of it's weaknesses. I have also tried the two more recently and have considered replacing the first one. However, instead of replacing, why not use them all and then compare the result? And that is exactly what I did.

ChatGPT

The first and the longest AI chat I knew in existence is ChatGPT. From the company OpenAI, they have released numerous versions as of this date of their GPT or Generative Pre-trained Transformer which refers to a family of models that OpenAI developed, popular for their advanced natural language processing capabilities. This is the capability of computer to understand our statements in english, allowing them to converse with human or human statements.

Advantages

  1. One of the advantages of this is it's age in the arena. It's been longer than any competitors, giving the creators of ChatGPT more learning from previous usages in developing newer and newer models. This is advantageous to the users because it will definitely improve the applications that they will be using.
  2. Aside from texts, users can also upload images or files if you are subscribed to their paid models. As of this writing, the GPT 3.5 is the free version they gave, so the latest models come with a fee or subscription plans.
  3. Another advantage is that there is a mobile app available so mobile users can access the app faster.

Disadvantages

  1. One of the disadvantage of ChatGPT is, in order to use a more advanced model, you have to have subscription. As of this writing, you need a GPT plus subscription which costs $20.00. 

Gemini

After sometime of the release of ChatGPT, Google announces their own large language model called Gemini. It's purpose is to compete with ChatGPT. And since almost all of us knows Google, it is a hope that more users will be using this tool rather than the above. However, it was never anticipated that being earlier of ChatGPT has taken its advantage.

Advantages

  1. The same with ChatGPT, Gemini can also analyze uploaded images without a paid subscription.
  2. As of this writing, for the free use of Gemini, we can already use 2.0 Flash model and 2.0 Flash thinking model with 1 million token window.
  3. The mobile app co-pilot is a great way to converse with AI much like you are talking with a person with a large set of data stored on its mind.

Disadvantages

  1. Before Gemini, Google introduces Bard, its AI conversational service and on the day of its demo, it produces an inaccurate information about the Webb Telescope. However, this has been long addressed by a series of evolution of the model and transforming it into now Gemini.

DeepSeek

As opposed to the two big players mentioned above, in terms of budget spent in the development, DeepSeek has the smallest cost in both development and training. With their own benchmarking, the results shows promise of competitive results on various subjects.

Advantages

  1. Based on the benchmark released by the company, they are exceeding in performance in the subject of coding. So if you want help with this area, you can try to give it a shot.

Disadvantages

  1. As of the time of this writing, deepseek experiences some server performance issues rendering it to respond slower than the 2 above. Not knowing when will this issue be addressed, this is the main bottleneck of the AI service. However, after a few seconds, the interface will still show the result and is still comparable with the results from the other 2 mentioned above.

Conclusion

Based on the experiences shown above, you have now a glimpse of what these AI services can do. Since today, we are now surrounded by things operated with the help of machine learning or what others might call AI. With these powerful tools, they can absolutely make our life better in so many things, however, with this power, we should also be very careful in its usage. Not everything they spits out are true, they are still based, in the simplest form, on a fitting of a curve, or predicting what will be the answer. Also there are issues about AI hallucinations where some services invents somethings that are not real. So we should be careful with these things and always carry out with the best of our judgement.

Thank you for reading! I hope you come back for more posts similar to these!

Labels:
Location: Floridablanca, Pampanga, Philippines

Comments

Post a Comment

Popular posts from this blog

Cursor AI Review: Is the AI Code Editor Worth It?

I've been using Cursor as my main code editor for a while now, and enough people have asked whether it's worth switching to that a proper review felt overdue. Short version: for me, yes — but with caveats. What is Cursor? Cursor is an AI-first code editor built as a fork of VS Code. That means every extension, theme, and keybinding you already use in VS Code works here, but with AI woven directly into the editing experience instead of bolted on as a plugin. It's made by Anysphere and can run models from OpenAI and Anthropic under the hood. What I like Tab completion is uncanny. Cursor predicts your next edit — not just the rest of the line, but the next change across the file. Once you get used to hitting Tab, going back to a plain editor feels slow. The Composer / Agent mode. You describe a change in plain language and it edits multiple files at once, showing you a diff to accept or reject. For refactors and boilerplate, this saves real time. It unde...
Post a Comment
Read more

MacBook Pro M5 vs M5 Pro: Which One Should You Actually Buy?

Apple's latest 14-inch MacBook Pro comes in two very different flavors: the base M5 and the step-up M5 Pro . On paper they look similar — same gorgeous Liquid Retina XDR display, same design — but under the hood the gap is bigger than the names suggest. Here's a clear, no-hype breakdown, with concrete use cases so you can match the chip to your work. Quick spec comparison Spec M5 M5 Pro CPU 10-core (4 performance + 6 efficiency) Up to 18-core (6 performance + 12 efficiency) GPU 10-core Up to 20-core Neural Engine 16-core 16-core Memory bandwidth 153 GB/s 307 GB/s (roughly double) Unified memory 16 / 24 / 32 GB 24 / 48 / 64 GB Max storage Up to 4 TB SSD Up to 8 TB SSD Battery (video playback) Up to 24 hours Up to 22 hours Media engines Single encode/ProRes engine More encode/ProRes engines (higher configs) What actually changes between them More cores — the M5 Pro nearly doubles CPU cores and adds GPU cores, so sustained, multi-threaded work finishe...
Post a Comment
Read more

How I used Google Sheets and Apps Script

Google Sheet is one of the most powerful spreadsheet application that exists online, rivaling with Microsoft's Excel. One of the main strengths is its strong support for collaboration with other users, much easier and popular than collaboration tools with Microsoft Office. Aside from plain spreadsheet, it also supports extensions such as macro. If you are familiar with macros on other office tools, they work almost the same. However, the most extension I use and tinker with is the Apps Scipt . Apps Script Extension One of the challenges I faced recently is how do I track or monitor reports in our department if they are submitted on time or worst, forgotten due to lack of better monitoring tools. So I thought if there can be simple applications that can be deployed or use by a more general user to allow reminding periodically what reports are approaching due dates or those that are past dues. Then I looked for a way, instead of creating a full blown app from scratc...
Post a Comment
Read more