LongCut logo

9 Gemini 3 Tricks That Will Help You Become a Pro

By Rick Mulready

Summary

Topics Covered

  • ChatGPT loses 12M users to Gemini
  • Prompt bigger with enhanced reasoning
  • Build apps from natural descriptions
  • Imagen 3 spells text correctly
  • Multi-step workflows yield pro designs

Full Transcript

Since Gemini 3 Pro launched a few weeks ago, at the time I'm recording this video, Chat GBT has lost 12 million daily users. That's a drop of about 6%.

daily users. That's a drop of about 6%.

And in the AI world, where Chat GBT has been the unquestioned default, that's a massive signal. And the reason for this

massive signal. And the reason for this is Gemini 3 Pro is the real deal. So, in

this video, I'm going to show you nine Gemini 3 tricks that will help you become a pro in no time. Let's get into it. Okay, so the first thing you need to

it. Okay, so the first thing you need to understand about Gemini 3 Pro is it fundamentally changes how much you can ask for in a single prompt. With most AI models, you've learned to be really

careful about prompt engineering, right?

You guide the AI through your exact process, giving it step-by-step instructions, right? Because if you

instructions, right? Because if you don't, you're going to get garbage outputs. And one of the big changes with

outputs. And one of the big changes with Gemini 3 Pro is its enhanced reasoning, meaning you don't necessarily need to have a super structured prompt. And what

that means is you can ask this model to do way more than what you're used to asking for from other AI model. You can

start to think bigger when you're asking an AI model to do something. And by the way, you get to access Gemini 3 Pro within the Gemini app. So that's what we're looking at right here in the screen. You just want to make sure that

screen. You just want to make sure that in the dropown here, you're choosing thinking because the fast option is Gemini 2.5 flash. Or you can access it

in Google's AI studio, which is a studio.google.com. google.com and you

studio.google.com. google.com and you can see it right here. So, we're going to use Google AI Studio for this first example here to show you how much we can

actually put into a prompt and ask Gemini 3 Pro to do. So, the goal here is I want Gemini 3 to help me to figure out

how we can generate 200 qualified leads per month in 90 days for this madeup company that I made up called Nimbus Ops, an AI workflow assistant for small businesses. So I give it context,

businesses. So I give it context, audience, positioning, inputs, give it the pricing, constraints, guardrails, and then I go through must haves. So I'm

asking it for strategy. I'm asking it to give me uh content drafts, landing page, email sequences, social media posts, offers, analytics, automation. [music]

So I'm asking it for a ton of things. So

I've just pasted that prompt into Google AI Studio. And check this out. In a

AI Studio. And check this out. In a

matter of this was like 2 minutes, it's given me an executive summary. It's

given me a strategy and decision matrix.

It gave me what it recommended. Gave me

a chart here. Here's the 90-day plan.

Messaging hierarchy, content drafts, email sequence, social posts, automation workflows, gives you the trigger. Um,

gives me the tools that I've told it that I like to work with, assumptions, and next steps. I mean, it just ripped this thing out in 64. So, just over one

minute. I could never ask a quote

minute. I could never ask a quote unquote normal LLM model for something like this because there's just so much here. And that's what I mean by thinking

here. And that's what I mean by thinking bigger when it [music] comes to prompting or asking Gemini 3 Pro for whatever it is that you want to do.

Another trick to understand is with Gemini 3 Pro also comes the ability here in Google AI Studio to set the thinking level. You can see it here. It defaults

level. You can see it here. It defaults

to high and then the other option is low. If you choose the drop down

low. If you choose the drop down thinking level controls how much internal reasoning if you will the model performs before responding. So if you switch to low it's usually going to

shorten the generation time and reduce sort of that verbose reasoning. The high

level is going to add some time add maybe seconds to the the model working for you but it's going to improve the reliability on more tricky prompts. If

your prompt is super simple you know low is not going to hurt. Now, before we jump into the next trick, I want to tell you about this tool I've been using for over a year now. We've all been there, right? You learn something valuable in a

right? You learn something valuable in a podcast or a webinar or YouTube video and then completely forget later on where you heard it. Recall is the AI powered knowledge base that fixes that.

You can use the browser extension to save, summarize, and chat with any content from YouTube to PDFs in just one click. It automatically organizes

click. It automatically organizes everything so that you can retrieve it later. And with their massive new

later. And with their massive new update, you can use their block style editor now to take notes right alongside all of your saved content. And the best part, you can chat with your entire

library of content. Ask Recall, for example, to pull insights from these business articles, these podcast episodes, these videos, and then combine them with my notes and draft a strategy

for me. To give Recall a try, go to

for me. To give Recall a try, go to getrecall.ai AI where you can try it for free or use code Rick 25 for 25% off the unlimited plan until January 1st, 2026.

The next trick with Gemini 3 is you can just bring your actual problem to this model, the whole thing, and it's going to reason its way through it just like a

PhD super super smart consultant would by asking the right questions internally in the model, evaluating trade-offs, and giving you a structured solution. And

what this means for you is you get to stop worrying about having the perfect prompt. So to show you what I mean here,

prompt. So to show you what I mean here, I'm going to paste in this really simple prompt. All I say is I'm drowning in

prompt. All I say is I'm drowning in admin work. I run a solar printer

admin work. I run a solar printer coaching business spending 15 hours a week just on email scheduling and invoicing. I use Notion, Gmail, and

invoicing. I use Notion, Gmail, and Slack. I want to automate as much of

Slack. I want to automate as much of this as possible. Propose a specific automation architecture. Reclaim 10

automation architecture. Reclaim 10 hours a week. Scenarios I should build first to get the biggest win. All right.

So in just a few seconds here, it's saying this is a very solvable problem.

Spending 15 hours a week because you're a coach. personal touch is your currency

a coach. personal touch is your currency blah blah blah. Phase one tells you what the bottleneck is, the proposed architecture, the workflow visualization, giving me the trigger. Uh

phase two broken out into phase three and then giving me a nice table here that I can export to sheets. So in a matter of seconds, it gave me a very helpful output here, a complete plan

broken down by different phases. And

this was the entire prompt. This next

capability is even wilder. I think it's about building entire applications. Even

if you're not a developer, you can describe what you want and it'll build for you a fully functional interactive app or tool. And so here are the instructions that I've given it here.

Build me a task manager web app called Sunrise Sunset. Totally just made this

Sunrise Sunset. Totally just made this up. The logic I'm telling it it needs to

up. The logic I'm telling it it needs to track tasks and projects. Super basic,

right? I give it a constraint. And my

constraint is just make it work. I don't

want to touch a config file or set up a database. Preload it with simulated data

database. Preload it with simulated data so I can see the interface immediately.

Uh and then finally for the tech part, give me a single HTML file with embedded CSS uh or JavaScript. Use advanced CSS for the lighting effects and make the

task burn away with an animation when completed. This is my entire prompt

completed. This is my entire prompt right here. So, it's building the app

right here. So, it's building the app right now. As you can see here in the

right now. As you can see here in the screen, it's coding it. Just a quick note here, I did choose the canvas feature. So, you just click on the

feature. So, you just click on the little uh menu here, and you can see canvas. And here is the here's the app.

canvas. And here is the here's the app.

Uh I'm laughing because this is so cool.

Uh remember I asked for the morning sunrise and the evening sunset. So, the

little uh basically a dark mode for it, which I think is super cool. Then it

gives all tides. What's the surf surf trip, grind, chill, buy tacos for dinner. Now, the other cool thing I want

dinner. Now, the other cool thing I want to show you here, this little trick, is click this little add Gemini features option. So, I can click this and it's

option. So, I can click this and it's going to give me some suggestions on the types of things it could add to it.

Okay, so it actually added two new features for me. It said that it added two new AI powered features to the app using the Gemini API to bring in some AI

to the beach beach vibes, catch waves, project suggestions, and the swell, which is the task refiner. So you can see here that added this catch waves button. Also this little AI icon here

button. Also this little AI icon here that we can also click on. And again, I know nothing about coding. I'm not a developer. If you said, Rick, you need

developer. If you said, Rick, you need to code this app. I would say this is what I'm this is exactly how I'm going to do it. So I could grab the code here and I could paste it into a page where I want to host this. Oh, and by the way, if you want to make any changes to the

app, let's just say I want to change the colors or what have you. This is where the vibe coding comes in. Meaning I can just ask it to make a change and it will it'll change it. Now the next trick with

Gemini 3 is understand that it's also improved in a big way with visual reasoning. For example, take a picture

reasoning. For example, take a picture of a sketch that you've drawn on a piece of paper. Upload it to Gemini and just

of paper. Upload it to Gemini and just simply say build this. It'll analyze the drawing. It'll understand the layout and

drawing. It'll understand the layout and the functionality of your drawing and it will generate the code just based on your drawing. You can use this site

your drawing. You can use this site right here. This is something that

right here. This is something that Google put together in Google AI Studio called Bring Anything to Life, which is exactly what we're talking about here.

You can just simply click on the upload and upload your picture and it will build it for you. Okay, for the next trick that you need to know with Gemini

3 Pro, you now have Nano Banana Pro, which is the much improved version of the Nano Banana image generation model.

You can see right here, I'm in the Gemini app. Nano Banana Pro is now on

Gemini app. Nano Banana Pro is now on Gemini 3 Pro. When it comes to prompting Nano Banana Pro, here are some prompt elements that you might want to make sure that you include. So, you want to make sure that you tell it what the

subject is, who or what is going to be in the image and be specific as possible. Composition. So, how do you

possible. Composition. So, how do you want the shot framed? For example, do you want it like ultra closeup or a wide shot or what have you? Next thing is give it action. Tell it what you want

happening in the image itself. Tell it

where you want this image to be. So give

it a location. Where does this scene take place? Also give it style like what

take place? Also give it style like what is your what's the aesthetic that you're looking for? And you can also use

looking for? And you can also use Nanoaden Pro for editing which is really really good at editing. So here's the prompt that I'm going to test this out with. And I'm going to give it the goal.

with. And I'm going to give it the goal.

So create a sunrise poster for a hipster coffee shop. I'm a huge coffee uh huge

coffee shop. I'm a huge coffee uh huge coffee fan. Love coffee. And so I want a

coffee fan. Love coffee. And so I want a coffee shop. I'm giving it the subject,

coffee shop. I'm giving it the subject, composition action location style camera and lighting, aspect ratio, etc. So again, I'm being very specific in my prompt. Now, once I have the prompt in

prompt. Now, once I have the prompt in here, I want to go down to my tools menu here and click on create images. That's

how you're going to be accessing the Nano Banana Pro model. So, I'm going to click on go. All right. And here is the image that it just created for me. So, I

asked it to make a flyer for this hypothetical coffee shop called Dawn Brew. and it added in sunrise specials

Brew. and it added in sunrise specials right here. A huge thing to know about

right here. A huge thing to know about the new Nano Banana Pro model because of Gemini 3 Pro is the spelling of words it gets right. Imagine that. So you can see

gets right. Imagine that. So you can see even back here on the blackboard on the on the wall it says single origin pourover. It has cortado has flat white.

pourover. It has cortado has flat white.

So it's spelling words correctly. Here's

another trick with Gemini 3 Pro. The

multimodal capabilities with Gemini 3 are really, really improved. Let's say

you have a podcast and you want to provide a list of timestamps on your show notes page. Well, you can upload the audio file to Gemini 3 Pro. It will

do that for you. It'll analyze the audio file. Now, I sold my podcast in 2023.

file. Now, I sold my podcast in 2023.

So, I'm just going to pick this other podcast that I really, really like. It's

called Planet M. I'm going to go to listen notes here and go to this uh episode, the neuroscience of success.

And I'm just going to go up here to the top and I'm going to click on download audio. So I've already done that. Then

audio. So I've already done that. Then

I'm going go over into Gemini and I'm going to upload the audio file. And I'm

just going to say listen to this audio fully understand the full podcast episode and then give me a list of the main takeaways with timestamps. So this

would be perfect for your show notes page. And so it gives me this great

page. And so it gives me this great overview of the uh of the episode. So it

gives me the intro uh with the timestamps. So secret number one. You

timestamps. So secret number one. You

could also ask it for the timestamp range, but then again gives you the uh secret number two, secret number three, and then ask me for a next step. So for

example, the host mentioned a specific exercise called confidence work around the 33 minute mark where you list out would you like me to help you start that right now? Or another trick you can do

right now? Or another trick you can do with this is Gemini 3 will analyze YouTube videos for you. So let's try it with this. So I'm going to say summarize

with this. So I'm going to say summarize this video and give me a list of the biggest takeaways along with how to implement each takeaway in an online business. So I've pasted the URL for the

business. So I've pasted the URL for the YouTube video in here and I'm going to click submit and let's see what happens.

Okay, that is a 46-inute video that it just processed in a matter of seconds.

And so gives the overview of what the video is about. gives the video summary, biggest takeaways, and then how to implement these takeaways within an online business. This is a super super

online business. This is a super super simple way to summarize this content and actually put it into and implement it in your business. Here's another idea for

your business. Here's another idea for you for YouTube videos and leveraging Gemini 3's multimodal capabilities is that you can use the model to get repurposing clip ideas from short form

videos. So, here's the prompt that I put

videos. So, here's the prompt that I put in. Analyze this YouTube video. give it

in. Analyze this YouTube video. give it

the link and it's a link of one of my videos and suggest to me short form videos social media clips ensure all clips are 30 seconds or less provide a list of topics with respective timestamp

range for each and I can click go okay and it's given me seven short form video clip ideas under 30 seconds so clip number one stop overpaying hook gives me

the topic why it works gives me the timestamp range and then here's the quote that is being mentioned in the clip it's pretty good uh gives me clip number two, clip number three, and it

goes all the way through for uh seven different clips. I can now very easily

different clips. I can now very easily give this to an editor if I wanted to to create clips from this this specific

video or I could do it myself uh with a tool like Opus clip or descript for example with the uh timestamp [music] range makes it so much easier. So for

this next one, Gemini 3 Pro is really really good at creating infographics. If

you caught last week's video I did here on the channel all about use cases with Notebook LM, which I'll link to right here, you saw the level of really cool infographics that you can now create inside of Notebook LM, but you can also

create them in here in Gemini 3. So I'm

going to say generate image of an infographic explaining the concepts presented in this video. I'm going to click go. Okay. And here is what it

click go. Okay. And here is what it created in just a couple of minutes.

here. It looks really busy, but it's there's a lot in that video. We can say, "Stop wasting money on more AI tools. Do

this instead." So, I did this in a matter of a few minutes here in Gemini 3. And for this final Gemini 3 Pro

3. And for this final Gemini 3 Pro trick, we're going to combine a few of the capabilities we've been talking about here today into one request into Gemini 3. So, what I want to have it do

Gemini 3. So, what I want to have it do is I want it to create me a new channel banner. So that's going to save me from

banner. So that's going to save me from hiring a designer or spend the time to do it myself, which I am not a designer in any way. And so this is going to go through multiple steps, right? It needs

to go to my YouTube channel, understand what the channel's about. I'm going to ask it to look at the thumbnails from the past few weeks on the channels cuz I want it in the same kind of look and

feel of those. And then I wanted to take that analysis and use Nano Banana Pro to create me a new YouTube channel banner.

I would like you to create me a new YouTube channel banner for my YouTube channel. Please go analyze my channel.

channel. Please go analyze my channel.

Here is the URL. Understand what the channel is all about and then look at my thumbnails from the past 3 weeks. You

should follow best practices for dimensions of YouTube channel banners.

All right, so there's the prompt I just threw in here. All right. Now, I just got to get the URL of the channel, and I'm going to choose create images, and let's see what happens here. All right.

And here's what it came up with. Now,

this is obviously is not me, so I would want to upload images of me for it to include, but it did what I asked it to do, and I think it did a really, really

good job. Now, this is pretty busy, so I

good job. Now, this is pretty busy, so I would essentially give it feedback and have it make edits to this image here.

But overall, I'm really happy with it how it handles this type of multi-step, multi-dimensional [music] if you will, workflow. So, there you have it. Nine tricks that will help you

have it. Nine tricks that will help you become a pro with the new Gemini 3. As

you can see here, it's really, really good. Just take one of these that you've

good. Just take one of these that you've learned here from the video today and go test it in the business. Just just pick one and go implement it and work your way through. Hopefully, this gave you a

way through. Hopefully, this gave you a lot of good ideas for your business. By

the way, if you're an online business owner and you're looking to leverage AI in your business, I want to invite you to join us inside my AI playbook community where I teach online businesses how to leverage AI so that

you can streamline your business, become more efficient, make it more profitable, and make more money in less time in your business. I'll link to it in the

business. I'll link to it in the description below. Thanks so much for

description below. Thanks so much for watching. As always, I'll see you in the

watching. As always, I'll see you in the next

Loading...

Loading video analysis...