Which AI is the Best? (Claude vs ChatGPT vs Gemini)

Anthropic, OpenAI, Google Gemini. How do

you choose one to commit to when there are so many good models out there? I

happen to have a pro subscription for all three. So, I thought it would be

all three. So, I thought it would be interesting to compare them in this video. If you're new here, please like

video. If you're new here, please like and subscribe. This is also part of a

and subscribe. This is also part of a 10-week series where I introduce people to using AI in their life and their business. So, check out more videos if

business. So, check out more videos if you're interested. This video is not

you're interested. This video is not going to talk about technical specifications. I'm not going to tell

specifications. I'm not going to tell you based on benchmarks which agent is best for coding because the problem with that is there's a new model every week and I think if you're subscribing to these and building up a pipeline for one

of these three systems you need to think long term and you need to stop thinking about what is the best model now and so in that framework I wanted to talk about a few different concepts feel free to jump between timestamps down below and

then at the end of the video I'll touch on grock and some other models briefly as well. So, first of all, from the

as well. So, first of all, from the cost, price performance, obviously prices are subject to change, but there isn't really a clear winner because between the three, they all have sort of

$5 to $10 beginner plan. You typically

have a $20 plus plan, and they all have sort of a $200 to $250 pro plan available. The names of the plans and

available. The names of the plans and the exact pricing change between the three, but it is very similar. And then

in terms of API costs, we'll jump into this a little more at the end as well, but it also is quite similar. they have

comparable pricing between the top three for different of their high tier models when it comes to calling the API rather than a monthly subscription. So, in

terms of that, it's hard to declare a clear winner, but I wanted to get the pricing out of the way first. In terms

of how you actually choose a model, I think there's a few important things to consider. And number one is what do you

consider. And number one is what do you actually have access to already and what are you using in your life, in your daily business? most of us because if

daily business? most of us because if you have something like Google Workspace for example, a small business running this, you already have access to all of Gemini's tools and you may not want to

also subscribe to OpenAI or Anthropic.

Similarly, if you're using Gmail and Google Calendar in all your life, the other tools do integrate well with them, but Gemini owned by Google as well is going to natively integrate all those tools, which means it's going to be

pretty futureproof and you're going to get access to things quicker than you might with the other tools. Similarly,

if you're a Microsoft Maxi, you might take a look at Copilot because they've baked it into a lot of their 360 subscriptions and you might not even want to look at the top three. Your data

is also something to consider and it's one where there's not really a clear winner, but out of the box, Claude Enthropic is the most quote unquote secure because they by default do not

use your data to train the model. For

most of the other systems, the basic plans do use your data in their training unless you get the enterprise level subscriptions, in which case they say they do not. But there are a lot of

security and speculations around data and privacy. And to be honest, you need

and privacy. And to be honest, you need to do separate research on that specifically because there isn't a clear winner when it comes to data privacy on which model you should choose. But if

you want me to dive more into that in a video, let me know down below. I think

really use cases is a good way to distinguish between the three. And I

actually for a fun exercise decided to ask all three models what they think the best use cases are for each of them. And

we'll go through that in a second. And

then after that I'll show you which ones I like to use for what use cases. So

this is the exact prompt that I gave all three. And I'm not going to read you

three. And I'm not going to read you everything here, but we'll go over briefly what they each said. So in one sentence per company, compare OpenAI, Anthropic, and Google Gemini in terms of their current strength and primary use

cases. Then list two to three specific

cases. Then list two to three specific tasks you would use for each. And I told it to search the internet to get the most recent data. This is from chat GDP 5.2. They say that OpenAI is best if you

5.2. They say that OpenAI is best if you want all-around assistance, especially strong for interactive do work with me type of problem solving. So they would use it with producing strategic docs end

to end things like debugging and refactoring code and long text context synthesis. They use anthropic. They say

synthesis. They use anthropic. They say

anthropic is best for high relatability agentic work. They would use it for

agentic work. They would use it for things like multi-step coding tasks like running an agent across a repo, long text content review/editing, large specs and contracts and computer use style

tasks. The agent thing is something

tasks. The agent thing is something we're going to see again in a minute.

That's a recurring theme. It seems like a lot of the models recommend Anthropic for using in conjunction with an agent.

If you haven't seen my Open Claw tips and tricks videos, go check that out. In

terms of Google Gemini, according to OpenAI, they say it's best for top tier reasoning and tight integration with Google's ecosystem, which is something I covered already. If you're using

covered already. If you're using Google's ecosystem, makes sense. They

would use it for complex reasoning problems and assistance with Gemini in Chrome. I don't really think this is a

Chrome. I don't really think this is a very good use case for Gemini in general except for the Google ecosystem. So

let's take a look at what Google actually had to say. This is how Google compares with the same task the three different companies. So they say that

different companies. So they say that OpenAI remains the leader in pure logic advanced mathematics and general purpose tool calling. They call anthropic the

tool calling. They call anthropic the solidified in a position of premier coding agent. So once again this is the

coding agent. So once again this is the coding agent use case coming up for anthropic. So that's recurring. They

anthropic. So that's recurring. They

emphasize nuanced reasoning and superior coding quality and safety first architecture. I don't know if I agree

architecture. I don't know if I agree with that with OpenAI having access to codecs is a very strong coding model, but this I do agree that I quite like using Opus 4.6 with OpenClaw and when you're coding with an agent, Anthropic

is a very good choice. Google dominates

in multimodal processing and massive context tasks. Now, this is something

context tasks. Now, this is something that you're going to see both Google and Gemini talk about Gemini and something that I agree with. It's very good at multimedia and it's very good at large context windows. So leveraging his 2

context windows. So leveraging his 2 million token window and deep integration into Google Workspace ecosystem to analyze entire repositories. So this is something that

repositories. So this is something that you can do with the other models, but sometimes they struggle if you're looking at big repositories. So finally,

what did Enthropic have to say about using all three of them? And you'll see some overlaps here, but basically they said AI is the consumer adoption. This

is true. OpenAI is currently leading with 800 million chat GPT users. This is

for just the chat. And OpenAI offers a broadest ecosystem and dominates industry benchmarks with strong all-around reasoning. It's best for

all-around reasoning. It's best for tasks like reasoning and business strategy, code generation, and prototyping plugins and ecosystems. Now, anthropic defines itself as being good

at high context APIs and enterprise contracts with 300,000 plus business customers. This is true. Even the

customers. This is true. Even the

government is using anthropic even though they might drop that if you haven't seen my latest short. And it's

good for once again they define themselves as being good at agentic coding complex debugging long document analysis and enterprise workflows. So

the enterprise and the agentic coding is something that comes up for a topic again and again. And then with Google once again multimodal tasks. So video

imaging and text analysis is something that Google is very good at. Massive

codebase analysis. This is once again if you have entire code bases in high context. The large token window of

context. The large token window of Google allows you to do that without having to break up sessions. So what do I do? I use OpenAI for deep research

I do? I use OpenAI for deep research because I quite like their deep research tool even though it's fairly limited even on the pro plan as well as my general ask Google. Ironically, I use

OpenAI when I have a quick question. I

find it gives me the best all-around answer. When I want something probably

answer. When I want something probably less biased, honestly, I might ask Claude, but I mostly use Claude for planning. If I'm going to do a product

planning. If I'm going to do a product launch, build an app, build a website, and I really want it planned out clearly with the steps. I really like Claude's artifacts. It lays out very nicely what

artifacts. It lays out very nicely what you need to build in what steps, and it can be very good for technical products or product launches. I also use it for OpenClaw when it comes to the API. And

when it comes to coding, I like Claude a lot. I mostly use Google and it's

lot. I mostly use Google and it's probably the one I use the least for if I want to fact check something I want to compare to the other models or when I want to do image generation I do use

Google even though OpenAI does have some good image generation tools as well especially through the API. If I had to choose one now that Codeex has been developed and is very good for coding I

would probably go with OpenAI as the all-around choice if I had to pick just one. That being said, I do use Google

one. That being said, I do use Google Workspace for business and so if I was you and I didn't want to pay for additional subscriptions, I would go to Google if you had access to it already.

And then let's touch on a few others briefly. Grock, for example, is great if

briefly. Grock, for example, is great if you want to do research on social media.

And because Grock is plugged directly into X, whether you're using the API or whether you're using Grock directly, it is great for scraping social media and finding what is trending compared to some of the other models. That's

primarily what I use it for. It's also

built in. If you're already paying for X, paying for Twitter, you have access to Gro. So, it's a great tool to use.

to Gro. So, it's a great tool to use.

And then you have some of the Chinese models, DeepSeek, Miniax. These are

great if you want a high performance at a lower cost. They tend to be cheaper than the top three and you do get comparable results, maybe because they're stealing some of the database

from Google and from Anthropic. That's

another topic. But I've also used their API for things like Kimmy from Moonshot, things like Miniax. When it comes to coding, they do a fantastic job and with your agent workflows with things like

OpenClaw. They are a great alternative

OpenClaw. They are a great alternative to be a lowcost solution if you don't want to pay for the tokens to use Opus 4.6 for example. So don't rule those out either. They also have subscriptions for

either. They also have subscriptions for chats. You can also check those out if

chats. You can also check those out if you want lowcost solutions for good integrations. I don't know if I would

integrations. I don't know if I would trust them with, for example, plugging into my Google calendar and Gmail, but that is up to you. Let me know down in the comments which model you're using

for what. Tell me why. And let me know,

for what. Tell me why. And let me know, is there something I didn't talk about you'd like me to cover in a future video? Thank you for watching. Have a

video? Thank you for watching. Have a

great day.