There are approximately 1.98 million public AI models as of 2025, with new ones being released almost every single day.

Yes, we all know that there are popular chatbots such as ChatGPT and Gemini, but whilst a few people are latching onto more advanced AI platforms, the vast majority have no idea what AI is truly capable of.

There is a lot to try, but this week I selected 10 AI models I believe will help make your business more productive and efficient.

You can watch me test each of these in my new YouTube video.

Fathom – The AI Notetaker

I had heard of an AI notetaker from multiple companies, but I didn’t really see the benefit – surely letting AI take your notes might make you stop paying attention in the meeting… how is that productive?

So I tried it for myself. Fathom joins as a meeting participant and can work across Google Meets, Teams, Zoom and more. It silently transcribes your entire meeting, and then summarises the notes into bullet points with relevant subheadings. In the app, you can also ask Fathom specific questions about the meeting, and it will pull information directly from the transcript to answer you.

For example, if you asked ‘What was the ROI for quarter one?’, Fathom would find this information instantly

My only qualm with Fathom was that it was visible to everyone in the meeting, rather than working discreetly on your computer. This may raise some questions, because I know firsthand that before this experiment, I had no idea why I was having meetings with people who had a ‘Notetaker’ on with them.

If you’re looking for a more discreet alternative (like I was), you should opt for Granola.

Granola – A Discreet AI Notetaker

Granola does not join your meetings; rather, it records directly from your device. Despite this, it can still record the entire transcript of your meeting and summarise it exactly as Fathom does.

The only feature, as far as I found, that Fathom has that Granola doesn’t is the AI Chatbot, where you can recall information. That said, I don’t know if that is too much of an issue unless your meeting is hours long, with multiple speakers and a lot of statistics to sort through. Both platforms concisely summarise, so sifting through the notes after the meeting shouldn’t take too long.

From the two AI Notetakers, I will be using Granola going forward. It removes the need to explain why ther’s a notetaker in the meeting, and avoids putting anybody under pressure.

OpusClip – Long Form into Short Form

OpusClip repurposes long-form content into short form. You simply paste in the link of a video or upload one from your device, and then it gets to work.

I was shocked by how quickly OpusClip produced 10 short-form videos. If I were to repurpose my YouTube videos myself, for example, it would likely take me 1-2 hours to produce 10 high-quality short-form videos from each.

Speaking of quality, I wasn’t sure how inclined I was to actually use the clips OpusClip produced. Whilst impressively resizing and moving elements, such as my face-cam to the bottom of the screen, and adding accurate captions, I wouldn’t say it was particularly good at selecting specific parts of the video.

NOW, that being said, I trialled this platform with two of my YouTube videos, neither of which were necessarily made with short-form in mind. Therefore, I will give OpusClip the benefit of the doubt and say that I may have made it a little trickier for it.

I can see this becoming a really helpful tool as it grows and gains more features. It was completely free to use, which I was also very impressed with. As a further test, I will repurpose this week’s YouTube video and see what traction it gains, then perhaps my mind will be changed.

Midjourney – AI Photo Generation

Midjourney is arguably one of, if not the, best photo generator on the market right now. Over 16 million people currently use it worldwide, with daily active users ranging between 1.2 and 1.5 million. 

I browsed the internet for examples of what Midjourney could produce, and I was blown away by the gorgeous creations that I found. So I decided to try it out for myself.

What is very important to note is that Midjourney requires an insanely detailed prompt in order to produce a high-quality image. Whilst this is the same across other AI platforms, there is an added emphasis for Midjourney. I used a Chatbot to expand and develop my initial prompt to make it most effective for Midjourney. Here is an example of a prompt I used:

“colossal futuristic megacity inspired by Tokyo urban density, endless layered districts, elevated transit rings, stacked neighborhoods, skybridges and vertical streets, gigantic civic megastructures rising above dense city blocks, massive scale contrast between tiny people and monumental architecture style: painterly cinematic illustration, precise linework with soft edges, ultra detailed but atmospheric color palette: deep ultramarine night sky, glowing amber and soft magenta lights, cool steel blues, warm window light scattering through haze lighting: dramatic volumetric light shafts, atmospheric perspective, luminous fog between towers, reflective rain-soaked surfaces mood: awe-inspiring, grand, epic civilization at peak advancement, hopeful not dystopian composition: wide panoramic vista from high vantage point, layered depth fading into horizon, sense of infinite city materials: matte metals, ceramic panels, glass but not mirrorlike, lived-in density”

Yeah… pretty long. But what was produced was absolutely stunning and was worth the extra time it took to ensure the prompt was spot on.

Midjourney and AI photo generation in general have many use cases. It can be used across advertising and marketing, as well as social media and blog posts. That said, I want to note that overuse of AI-generated content can damage your brand’s reputation and should be used sparingly as a tool rather than a mechanism.

This is the same for copy generation. Creating a website or blog post entirely with AI will negatively impact your brand image. I promise, people will notice.

Now, with that disclaimer out of the way, I couldn’t possibly say that Midjourney wasn’t impressive. It created gorgeous pieces very quickly. At this stage in my business, I don’t see AI photo generation as very useful, but I can see how it could be helpful for some.

But if you read one thing in this article, let it be this. AI IS A TOOL, NOT A MECHANISM.

VEO3 – AI Video Generation

VEO3 is undeniably, insanely impressive. In less than two years, VEO has progressed to now making hyper-realistic videos. If you’d like to watch a few examples, click here. 

I tried it for myself. There were a few flaws, but perhaps they stemmed from the prompts I used. Like Midjourney, it requires a lot of detail to achieve the desired result.

VEO3 was hard to get my hands on. You can get it through multiple platforms, such as Artlist or Google (its creator), but this can be expensive. I used a free trial in order to access it, but I can only produce videos up to 8 seconds long.

Possibly the most impressive feature of VEO3 is the sound. It created personalised sound effects that perfectly matched what happened in the video produced, and even created voice-overs for certain characters. If you would like to see what I created with VEO3, you can watch my YouTube video <here>.

Like Midjourney, AI video creation has many uses, including marketing and branding. However, I put even more emphasis on the fact that I don’t support it taking over industries, and I want to make clear that there should be a priority for video editors, graphic designers, and everyone else in the creative industry who worked hard to get here.

That said, AI video creation will only continue to improve, so I was amazed to see its progress just 18 months into its production.

ElevenLabs – AI Voice Generation

ElevenLabs blew me away when I first discovered it. You can select from an extensive library of voices, or even create your own.

You simply type in what you want to say, select a voice, and click generate. Your sample is produced in seconds.

Their newest model, Eleven V3, produces two samples every time you produce something. You can also use square brackets ([ ]) to make changes to your voice, such as:

• Intonation
• Tone
• Mood
• Sound effects (e.g. [laugh])

This creates hyperrealistic results that can be fine-tuned to perfection for what you need.

This can be very helpful, and has been something I have utilised before this experiment. Having the choice of lots of voices means you can get voiceovers for all of your use cases, and do so in the snap of your fingers.

Furthermore, voices are created from real voice samples and creators, who are compensated each time a voice is used. This means that all voices sound natural, because they’re made by real people!

Apollo & Instantly – Prospecting Databases

Apollo and Instantly are prospecting databases where you can find leads for your business, and set up automations in which to reach out to them.

I found that Apollo was much more technical, with many intricacies that allowed for a very efficient, specific workflow. Instantly seemed more basic, yet it had the same core functionalities I found in Apollo. It is common for both to be used together to create the most efficient workflow.

I know from speaking with people in the industry that the use of these platforms is becoming increasingly popular to scale outreach and send up to hundreds of thousands of emails a month. This can obviously boost productivity and output significantly, but it is vital that it is set up effectively to achieve the best results.

Whilst AI platforms should be human-led, their input should be human-led too, so outreach is personalised and human, not sounding clearly AI. This will have the greatest impact on results.

Because of the price, though, it is popular to use other platforms that are cheaper (or free) in order to scrape lead lists.

Gamma – AI Presentations

I was amazed by Gamma. You simply paste in a large chunk of text, and it will create a gorgeous presentation using relevant images, icons and colours.

You can select from a large library of different themes, or even make your own that uses your brand colours.

Watching it work was incredible. It took under 30 seconds to create a 10-page presentation, with all of the necessary information. Then, with a single click, you can export for free. Whilst there are paid plans, you get 400 credits upon signing up, which is enough for about 10 presentations.

Like with all AI, you will likely need to make some edits, either to spelling, grammar, accuracy or the god-foresaken AI dash (-). After that, though, you will have a stunning presentation that you can use for social media or real-life presentations.

To get the best results, ensure the text you’re entering is high-quality. Gamma can produce a presentation from scratch, but whilst efficient, this is unlikely to produce the most accurate results for your use case.

Jasper – AI Content Assistant

Jasper offers 100 apps you can use, from video scriptwriting to blog writing to photo generation.

I was slightly spoilt for choice, but I chose to trial the video script and blog post. It worked quickly, but definitely had inaccuracies, one of which was thinking we were still in 2024, and another was not capitalising words at the beginning of a sentence.

I couldn’t help but think that Jasper didn’t seemingly offer anything that I couldn’t get with a regular chatbot. For $69/month, I struggled to find what differentiated it from the other platforms on the market.

I wouldn’t jump to use Jasper again, but I can see how it is a strong hub for everything a marketing agency or creator needs for day-to-day work.

Final Thoughts

I can proudly crown (drum roll please)… GAMMA is the winner of this challenge.

Gamma produced genuinely impressive work, and the biggest advantage for me was the time it saved. This can definitely boost a business’s productivity, and I urge you to try it for yourself.

Granola was a close second, and is definitely an app that I will be using again.

I’d be interested to hear which is your favourite! You can let me know by emailing me here.

Want to know which AI model will help YOUR business? Book a 30-minute AI productivity audit, and I’ll map the right stack for your workflows.

TL;DR

1. Fathom – AI Notetaker: joins as a meeting participant, which may be an awkward one to explain
2. Granola – AI Notetaker: works only on your device, making it more discreet
3. OpusClip – Repurposes Content: great captions and resizing skills, queries about varility potential of the clips produced
4. Midjourney – AI Photo Generation: impressive, high-quality images, but requires immensely detailed prompts.
5. VEO3 – AI Video Generation: stunning visuals with insanely impressive sound, but limitations with length and accuracy
6. ElevenLabs – AI Voice Generation: versatile and can be tailored to very specific requirements
7. Apollo & Instantly – Prospecting Databases: can be effective if set up properly. Price means that people use cheaper alternatives. Method is seeing strong growth in popularity
8. Gamma – AI presentations: incredible quality and works very efficiently, cheap pricing and great visuals to be used as you require
9. Jasper – AI Content Assistant: no features that differentiate it from a regular chatbot