GPT-6 Astra vs Fable 5.1: Eight Real Demos, One Verdict
A host runs eight creators' AI-agent experiments back to back, from a $60K app clone to a 150GB video edit, and scores which model actually wins each job.
Two flagship AI agents now build working software from a single prompt, but the real gap between them shows up in computer control and cost, not in raw output quality, which is close to a tie.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
You're deciding which AI agent to standardize on for building apps, editing video, or automating your computer.
You want real cost and time numbers for agent tasks instead of marketing claims.
You're curious what browser control, iPhone control, or MCP-based tool control looks like in practice before trying it yourself.
You run meetings or a small team and want to see what an AI agent finds when it reads dozens of your call transcripts.
SKIP IF…
You've already run both models side by side on your own tasks and just want the raw demo links, not a synthesized recap.
You have no interest in agentic computer/browser control and only care about chat-based writing or research quality.
TL;DR
The full version, fast.
A host distilled hours of demos from eight different creators into a head-to-head between OpenAI's GPT-6 Astra and Fable 5.1. Both models can build a working app, a video recap, a meeting analysis, or a consulting deck from a single prompt, and the visual output quality is close enough that reviewers split on which they preferred. The real separation shows up in three places: computer and browser control, where Astra took over a live desktop, an iPhone via mirroring, and DaVinci Resolve through its native MCP server with almost no manual clicking; writing, where a professional writer called Astra's prose crisp and free of AI-isms; and cost, where Astra ran roughly half the price and noticeably faster across a meeting-analysis task ($5.50 vs $46) and a consulting-deck task ($12.14 vs $26.61, in 23 minutes vs 37). The practical conclusion: pick based on the job, lean Astra when the task involves controlling software rather than just generating content, and always check the time-and-cost readout before assuming one model is obviously better.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
The host frames the video as a synthesis of hours of Astra demos from other creators, promising case studies, links, and his own verdict.
00:27 – 01:39
02 · Raycast-style app
A creator prompts Astra to build a Raycast-style clipboard/launcher app with image search and background removal, finished in under 20 minutes.
01:39 – 02:51
03 · The $60K app clone setup
Jason Lee sets out to clone a receipt-scanning bookkeeping app making $60K/month, giving the agent only the App Store URL, a Claude API requirement, and Notion-style branding direction.
02:51 – 04:12
04 · GPT-6 Astra's app build
Astra's version uses a clean green-and-white palette, working navigation, and a functional receipt scanner with haptic feedback; the host prefers it on design despite missing dummy data.
04:12 – 06:18
05 · Fable 5.1's app build
Fable's version ships with populated dummy data and correctly extracts and categorizes a real scanned receipt via a connected Claude API key, but the host finds the design less polished.
06:18 – 07:39
06 · Computer use in Chrome
Claire Vo has Astra take over her live Chrome browser to rebuild a complex node-based automation workflow, adding logic and reconnecting nodes hands-off.
07:39 – 08:42
07 · Sponsor: Zapier MCP
Sponsor break explaining Zapier MCP, which connects thousands of tools to AI agents while letting teams restrict what those agents can actually do.
08:42 – 10:12
08 · Browser control speed
Matthew Berman has Astra draw a research workflow in Excalidraw, compare rare Pokemon card prices, and plan a full Kyoto walking tour, each task finishing in under two minutes.
10:12 – 12:00
09 · iPhone control
Mark Kashef has the agent drive his iPhone through Mac's iPhone Mirroring feature, opening Instagram, verifying the account, and adjusting media-quality settings.
12:00 – 15:18
10 · DaVinci Resolve via MCP
A creator connects Astra directly to DaVinci Resolve's native MCP server and has it perform a rough video edit, placing footage on separate tracks, without touching a mouse; the host calls this the standout moment of the video.
15:18 – 19:12
11 · 150GB video recap
Given 150GB of raw event footage, both Fable 5.1 and Astra generate a short, music-scored recap video; the host finds both usable but with rough spots, giving Astra a slight edge.
19:12 – 22:12
12 · Meeting analysis
The same set of leadership meetings is analyzed by both models for pain points and automation ideas; Astra reviews more meetings (79 vs 58) for a fraction of the cost ($5.50 vs $46).
22:12 – 24:09
13 · AI writing
A professional software-review writer has Astra draft its own self-critical 'vibe check' review, calling the output crisp and free of typical AI writing tells.
24:09 – 27:27
14 · Consulting decks
Both models generate a McKinsey-style branded consulting deck from the same prompt; the host prefers Fable's output but notes Astra finished faster and for less than half the cost.
27:27 – 28:03
15 · Final verdict
The host's overall take: Astra leads decisively on computer/browser control and writing, both models still need design guidance, and Astra is consistently cheaper and faster.
Atomic Insights
Lines worth screenshotting.
One reviewer built a working desktop utility app, modeled on Raycast, from a single prompt in under 20 minutes.
A $60,000-a-month receipt-scanning app was rebuilt from scratch by giving an agent nothing but the App Store URL and a request to use Claude's API for receipt extraction.
Astra's version of a cloned bookkeeping app used a clean white-and-green palette and was preferred on design by the reviewer, even without any design reference given in the prompt.
Fable 5.1's version of the same app used more color and dummy data but was judged less premium in layout and navigation icons.
Astra took over a live Chrome browser and rebuilt a multi-node automation workflow, adding new logic and reconnecting nodes with zero manual clicks.
Astra completed a multi-step Google Maps itinerary and a comparison of collectible card prices in under two minutes each, screen-recording its own work as it went.
An agent controlled a real iPhone by driving Mac's built-in iPhone Mirroring feature, opening Instagram, checking the active account, and toggling media-quality settings without any custom integration.
DaVinci Resolve ships a native MCP server, so an agent can perform a rough video edit, placing footage on separate tracks, entirely through structured commands instead of simulated mouse clicks.
Controlling software through an MCP server is faster and less brittle than controlling it by moving a simulated cursor, because MCP is a first-party integration rather than a workaround.
Given the same 150GB of raw event footage, both models produced a usable 60-second highlight recap with music, but reviewers still preferred hand-tuning the result over accepting either output as final.
Analyzing dozens of internal meeting transcripts, Astra reviewed 79 meetings for $5.50 while Fable 5.1 reviewed 58 meetings for $46, roughly eight times the cost for less coverage.
A professional writer who runs a paid software-review newsletter called Astra's prose crisp, to the point, and free of typical AI writing patterns.
On a from-scratch McKinsey-style consulting deck, reviewers preferred Fable 5.1's output for presentation quality, but Astra finished in 23 minutes for $12.14 versus Fable's 37 minutes for $26.61.
Across every test in the video, the two models' raw output quality landed close enough that different reviewers picked different winners on the same task.
Takeaway
Astra wins on control and cost, not on raw taste
WHAT TO LEARN
Two AI agents can now build an app, edit a video, or write a deck from one prompt, so the real decision comes down to who controls software better and who costs less to run.
02Raycast-style app
A single, well-specified prompt (a reference URL, an API to use, one design direction) is enough to get a working app scaffold in under 20 minutes.
When you don't give an agent a design reference, expect it to default to its own house style rather than something tailored to your brand.
03The $60K app clone setup
Cloning a proven, revenue-generating app is a fast way to validate that an agent can produce a real, functional product, not just a demo screen.
04GPT-6 Astra's app build
A clean color palette and consistent navigation can outweigh populated sample data when judging a freshly generated app's design quality.
Small interaction details, like haptic feedback on tap, are the kind of polish that separates a good one-shot build from a merely functional one.
05Fable 5.1's app build
The same one-shot generation gap shows up everywhere: giving no design direction gets you a functional app, not a distinctive one, regardless of which model builds it.
A model correctly extracting and categorizing data from a real scanned document, not just a screenshot, is the functional bar worth testing before trusting an app build.
06Computer use in Chrome
An agent that can see and rebuild a live, complex tool (not just a simple form) hands-off is a meaningfully different capability than basic task automation.
Watching an agent take over your own browser session, rather than a sandboxed demo, is the clearest way to judge whether computer use is trustworthy enough for your workflow.
08Browser control speed
Multi-step browser research tasks (comparisons, itinerary planning) that used to take a human many minutes now finish in under two minutes end to end.
An agent that screen-records its own work as it completes a task gives you a built-in audit trail for free.
09iPhone control
Built-in OS features like iPhone Mirroring extend what an agent can control well beyond apps on your computer, including your phone.
Before granting an agent broad computer or browser access, restrict what it can actually touch, especially for tools connected to real accounts and data.
10DaVinci Resolve via MCP
Native, first-party integrations (MCP servers) control software faster and more reliably than an agent simulating mouse clicks, because the app exposes its own commands directly.
If a tool you use already ships an MCP server, connecting an agent to it directly is worth trying before reaching for screen/computer control.
11150GB video recap
For a first pass on a large raw footage dump, an AI recap can get you a usable rough cut, but treat it as a draft to hand-edit, not a final deliverable.
When you need a quick, good-enough edit rather than a polished promo, a plain-language request to a general AI assistant may beat a dedicated agentic workflow.
12Meeting analysis
Feeding an agent your own meeting transcripts and asking for the top pain points and one automation idea is a low-effort way to surface real process gaps.
When comparing AI outputs for a real decision, always check total cost and time next to quality. A cheaper, faster result that's only slightly behind on polish may be the better call.
13AI writing
AI writing quality is judged by absence of tells (rhythm, hedging, filler) as much as by presence of good ideas. Read a sample out loud before trusting a model for client-facing writing.
Having a model review and critique its own creator's writing style is a fast, if biased, way to sanity-check tone before publishing.
14Consulting decks
One model asking clarifying questions before generating a deliverable, versus another running immediately, is itself a signal worth weighing when picking a tool for ambiguous tasks.
A branded, structured deck can still fail as a live presentation if it's too text-dense. Judge AI-built decks separately for 'send as a document' versus 'present out loud.'
15Final verdict
The model that wins on any single task can lose the next one. Test each of your actual use cases rather than picking one model to use for everything.
When two models are close on quality, cost and speed become the real tiebreaker, especially for tasks you'll repeat often.
Glossary
Terms worth knowing.
MCP (Model Context Protocol)
A standard that lets an AI agent connect directly to an application's internal commands, instead of controlling it by simulating mouse clicks and keystrokes.
Computer use / browser use
An AI agent capability that lets it see and control a real desktop, browser, or phone screen directly, clicking, typing, and navigating like a person would.
iPhone Mirroring
A built-in Mac feature that displays and lets you control a connected iPhone's screen from the Mac desktop, which an AI agent can drive the same way a person would.
Vibe check
An informal first-impression review of a newly launched piece of software, focused on look, feel, and immediate usability rather than a full technical breakdown.
One-shot
Producing a finished result from a single prompt, with no follow-up corrections or iteration.
“YouTube is blowing up with how amazing Astra is.”
cold-open hook stating the video's premise→ TikTok hook↗ Tweet quote
13:31
“It's blotted out the screen share. It's focusing on my face. Now look at this. This is happening in real time. Hands off this. It's really impressive. This is the best bit of the video.”
the reaction moment the host calls the video's peak, tied to a visible live demo→ IG reel cold open↗ Tweet quote
24:09
“Astra was five and a half bucks and Fable 5.1 here was 46 bucks, which just makes no sense to me.”
sharp, specific cost stat with an emotional reaction→ newsletter pull-quote↗ Tweet quote
24:22
“I do need something that conserves my tokens.”
relatable cost-conscious line for anyone paying for AI usage→ TikTok hook↗ Tweet quote
24:03
“Its sentences are crisp. They're to the point. There's no AI-isms. There's no slop.”
concrete praise for AI writing quality, quotable as a review line→ newsletter pull-quote↗ Tweet quote
27:11
“Fable was 37 minutes and Astra was 23 minutes, and Fable was 26 bucks and Astra was 12 bucks, so Astra was a little less than half the cost as well as 14 minutes faster.”
closing head-to-head stat that anchors the entire video's verdict→ IG reel cold open↗ Tweet quote
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
analogy
YouTube is blowing up with how amazing Astra is. He had it copy an app that makes $60 ,000 a month. She had it control her computer.
He had it edit a video and he lost his freaking mind about it. It's focusing on my face. Now look at this.
This is happening in real time, hands off. This is really impressive. I watched hours of video.
I used Astra myself. I'll show you the best use cases with links below to the creators and chapter markings so you can skip to the case studies you need. And I'll tell you what I think.
Let's get into it. Presented by Zapier, the AI automation company. I had this idea for an app that would look a lot like Raycast.
So Raycast is my clipboard manager and it's my app finder. It's a very cool app. I love how it just kind of pops up when you need it.
And so I just went to Codex, right? I just said, I want to create a desktop app that I can use via shortcut. And it's a Raycast style app that lets me have Google images.
and remove the background and all of them that I copy or save get saved to the saves folder. And then I just like had a bunch of ideas for the app. And then literally after 16 minutes, it created this app right here.
Command shift space. I can open this up and I can type codex logo. And notice here, if I press enter, it searches for my.
The Codex logo. And I could copy this right here and we could paste this in whatever it is that I'm working on. Here's the Codex logo.
In less than 20 minutes, he had the app that he needed. I'm finding more and more of us are using both, frankly, Astra and Fable 5 and all these tools to create apps fairly quickly. Let's go a little bit more in depth and let's take a look at this one.
And this is an app that's making $60 ,000 a month. So I want to replicate this app. I do like when people are replicating apps that make money.
The idea is maybe I can make money the same way, but obviously there's more to a business in the software. Having said that, I think it's important to see how he did it. We'll scan your physical receipts, detect all the items in the receipt and categorize everything and even connect to your accounting software.
So it basically automates manual bookkeeping. Okay, let's see the prompt that he used to do this. But at a high level, what I gave it was this URL.
of the App Store. So it knows what features to build here. I also said that I want to use the Cloud API for receipt extraction, meaning it needs to connect to some sort of LLM to be able to scan the receipt and interpret the data.
So I'm going to be using Cloud API for that. And the only design direction I gave it was to use Notion style branding for illustrations and design direction. And for everything else, I ask it to just be creative and give me the best shot.
And one more thing that I ask it to do is actually to produce the app screenshots as well. So if I want to publish this to the App Store, I don't have to recreate all this manually. Okay.
The cool thing about what Jason Lee did is he had it built using OpenAI's Astra and Fable. Let's take a look at both. First, OpenAI.
It created as well. And again, it defaulted to this green color palette, which I don't think it looks bad at all, actually. And I think overall, it looks pretty clean.
It doesn't have any of the dummy data that I asked it to populate the app with, but that's fine. So if I go to receipts, there's nothing here. But if you look at the way everything is placed, all the buttons, the colors, everything is just clean and premium.
If I go to reports or settings here, yeah, this is really good, actually. I'm pretty surprised. I would say from a design perspective, I would pick this one over Fable.
But let's check the functionality of the receipt scanner here. Okay, so I can choose from photos. Or take a photo.
Let's take a photo. And let's use photo. There you go.
So now I can scroll down and confirm all the numbers here. Also select the same category. Can change any of these line items here.
And then reviewed and ready. I can save receipt. And then if I go to all.
Yeah, there you go. So the receipt is here. So yeah, I think in terms of design.
Considering that it's just one prompt, I do like this one better. Even the icons for the navigation tabs down below here are actually pretty good. And it gives you this haptic feedback when I click on it as well.
Okay, it's nice. Like I said, he did this in Fable in addition to Astro. So let's see what he did with Fable phone.
And I'm going to start with what Fable built for us. So I'm going to click on receipt scan here, loading up the app. All right.
So this is the app that I was able to build. So it's got all this dummy data here that we can probably scroll through. We've got the receipts, reports, and we've got settings.
And let me try to actually scan a physical receipt here. So I'm going to click on camera icon here. I'm just going to scan this receipt that I have here.
Take a picture of that. All right. So it was able to capture that data.
And because I've given it my Cloud API key, it's able to use Cloud. to extract all this data and categorize everything correctly here. And now I'm able to actually pick a category here, I think.
If this is food or drink, okay. So I can click on groceries if I want to. But this is food and drink.
And I can edit any of these fields here. And I can even change the currency, I think. Yeah.
So I can change the currency. I can see if it's got all the names of the items correctly and the quantity as well. And yeah, looks right.
Hit save. and then it's saved to the database. Now let's look for that particular receipt that I just added.
If we go to receipts, the filtering here is a bit broken, which can be easily fixed, but let's go to food and drink. Let's look for, okay, so this one, yeah, there you go. Now looking at the design here, I think this is nothing special.
Although it is functional, it has all the features correct, but I wouldn't say that it's amazing in terms of design. But to be fair, I also did not give it any design reference. Okay, so the OpenAI version he liked better.
But honestly, if you take a look at both of these, when it comes to developing, to building a product, what I saw throughout all the videos that I saw is essentially they were neck and neck. Fable, Astra, some people liked one. Sometimes it came out that the other did better.
It's not enough for you to say, okay, this is what makes Astra amazing. You're going to see what makes it amazing in a moment. Let's keep going.
Here is Hawaii AI, Claire Vo. She is using it on... A tool that she has on her computer, on her browser.
And I'm going to kind of cut through the prompt that she gave towards the end is where we'll start. Let's get to it. Can you update the workflow in Chrome to this?
Thanks. OK, so what we should see Codex start to do is actually open up Chrome, take over Chrome and start building out new nodes, new logic, new prompts. and route all of this stuff, reconnecting the right nodes.
And here it goes. So you can see it's adding a new object to the AI generation that we're doing, who the email is going to, what the email subject is, what the email body is. I'm guessing, look, there's no hands.
I'm not doing any of this. And it's going ahead and going in this like fairly complex node -based app. figuring out where to make the changes, include things.
And I think ideally it's going to show an email draft from Claire or from Zach to review and for us to ultimately send. And so that's going to send to our Slack channel and then we can copy and paste that and email it to the prospect. This is the incredible part.
Computer use, browser use. This is where it really shines. And yes, it could do it before previous models could.
But this is so much faster, so much smarter. Let me show you how fast it is. Oh, but first, I got to tell you about Zapier MCP.
If you are using... Your tools, Notion, email, all these tools that you love and you're connecting it to agents, you're kind of asking for trouble. And even if you're great at doing it, maybe your team is doing it and they are asking for trouble by giving their AI too much power.
What Notion MCP does is it allows you to bring thousands of tools, literally thousands of tools into the AI tools that you want to use. but it lets you restrict what those AI tools can do with the software that has your precious data in it. It makes it easy to connect apps, but also easy for you to restrict the things that your AI does with it.
You can go try it right now at zapier .com slash MCP, and you will see why so many people who are at the cutting edge need the kind of protection and access that Zapier MCP gives you. I don't have, like, my own personal URL. I'd love it if you told me that you heard about them from Andrew or the next new thing.
Let's get into it. I was about to show you. This is Matthew Berman.
Here's what he did. I want to show you a few examples of how good Astra is at browser control. Check this out.
So here I had it open Excalibur and draw a research workflow. You can see the timestamp right here. It's going to skip ahead in a few parts, but you'll see the overall duration of time that it took to do this.
So check this out. Here we go. It's already adding text.
It's adding bubbles. It's at 17 seconds right now. Look at this.
31 seconds to do this. And there we go. completed in about 30 seconds now it's doing research on rare pokemon cards 30 seconds in 40 seconds in and i mean all of this gets done in under one or two minutes now it's running comparisons it's not just looking at the page so here we go that finished in a minute 38 seconds to look at all of these different cards compare them and now We're going to plan our walkthrough Kyoto.
So this is actually it using Google Maps. And by the way, it put together this entire video you're looking at. It recorded its own screen, put together the information on the left side.
And there we go. A minute 23 to plan an entire walkthrough Kyoto. Now, this is this is the stuff that I want you to see the speed that it works at.
And what I love about seeing use cases like him saying that he had open eyes tool. take a screen share for him is I had no idea you could do that. I need to see what other people are doing with this to see what I could do to give me ideas for what to do.
All right. So he just did three things in the browser. There's so much more I want you to see.
Let's take a look at this. He decided this is Mark Kashif. He said, I wanted to control my phone, which I had a hint that you could do, but I wanted to see it in action.
Here's how he did it. Check it out. Now that I've sent this over, it should open the iPhone mirroring app.
Go and search. By the way, if you are not on a Mac, you should know that, frankly, even people on Macs don't seem to know this, that there is a screen mirroring app that comes with every Mac that lets you see your iPhone and control your iPhone on your computer. And that's what he is using.
He's saying, hey, look, if OpenAI's tool, whether you call it ChatGPT or Codex, doesn't matter to me, if it can do something on my computer, then this iPhone mirroring app is on my computer. Let's have it control that. And more specifically, he's using it for controlling Instagram.
Okay. Enough of me. I want you to see how he did it.
Now that I've sent this over, it should open the iPhone mirroring app, go and search for the Instagram app, go through, double check it's on the right account, then go through the settings. So it's connecting to my phone. It should go and click on the search bar right now and enter the Instagram app.
There we go. It's clicking through. It spotted it without even searching for it.
It's already in settings and activities. So it's double checking that it's on the brand new right account. Then it's going to go back.
click through and do what it needs to do to make sure that everything is optimized as possible. So it goes to the media quality tab where it toggled on the highest quality. And it's just going through what seems to be any core things in its training data that tells it these are the core things to focus on.
Incredible, right? Okay. And I like the speed, the snappiness and the intelligence that goes into it.
This is what really makes Astra so incredible. Okay, this is the guy who lost his mind. And I totally get it.
You can have computer control, but sometimes computer control is too slow, you know, especially when you're doing something like video editing. And so in Creator Magic, when you want to do video editing, he said, I'm not going to have it click around a mouse like a human being. I'm going to have it connect using an MCP server.
And DaVinci, the video editing app that happens to be free, so way cheaper than the other. Then the others also has an MCP. Let's see how he sets it up.
Let's see how he prompts. And then I want you to see the actual edit. Check it out.
Ask GPT -6 Astra. Okay, I've got DaVinci Resolve running 21 .1. They say that it can integrate with AI agents via MCP.
Are you able to find that functionality and hook it up so that we can control DaVinci Resolve, not with a cursor, but with the MCP server? So I'm going to get Astra on the case for that and see if it can hook this up itself. Controlling a video edited by the cursor is not what I want to do.
And the reason for it is that controlling by cursor is pretty slow and janky, whereas MCP is like a first party thing. So it's checking the support right now in my local setup and connecting. okay it's found it your resolve 21 .1 installation includes blackmagic's native resolve mcp this is awesome checking the live connection and registering with codex okay you see it's as easy That is easy.
All right. Let's see the prompt and then really where the amazing part is in is in watching what it does. But I want you to see the prompt.
So can you see my open DaVinci Resolve session? Are you able to insert both the program and the screen share onto separate video tracks and perform a rough edit for me using only the DaVinci Resolve MCP, please? Now, that's quite a complex first initial prompt.
You'll see I'm using GPT -6 Astra High, and it says I'll inspect the open project. So hopefully it's going to see this OpenClaw2 MCP project and edit through Resolves MCP. I'm very specific here.
I don't want it to take over the computer. I don't want computer use. I want this all done by the MCP link to see how good it is.
Okay, hands off at the moment. It's thinking things through. It's made a change to a file in the project.
We'll see what comes back. Now it seems to be running a lot of scripting here. A lot of Python is getting created.
Resolve MCP client Python. OK, just for orientation, maybe I should have said this before. On the right, he just has the Codex app from OpenAI.
And on the left, he has his video editing tool. Now let's see the actual edit. You saw the prompt.
You saw how he connected it. Let's see what it does. What's happening here?
It's blotted out the screen share. It's focusing on my face. Now look at this.
This is happening in real time. Hands off this. It's really impressive.
This is the best bit of the video. Oh, my goodness me. Okay, he is the most excited that I've seen about this, but I have seen other people use this.
The ability to connect through MCP is even faster than through computer connection, but they're both really fast. And this is really where Astra is amazing. And I wonder how many people are actually using it.
If you are, let me know in the chat. I'm curious about whether this is a tool that people are actually using day to day. I know I am using it a lot.
Okay, so what if you don't want to actually see the micro edits and go back in into DaVinci Resolve and edit yourself? Well, you can just give it a... bunch of video files and have it give you an output.
I did that. All right. This is Nate Hurk.
I gave it this folder, which has over 150 gigabytes of resources and recordings from our AIS live event in July. And I asked them to make us basically a recap, like an energetic, real 60 seconds that shows like a recap of the events that we can use to promote the next ones and things like that. Now, I'm really impressed with both of these outputs.
Let me just say this in favor and ask how much information it had to look through. And. how it had to create a story and figure out how to put everything together with music i think they both did really really well so here is fable 5 .1 hello hello ais live man let's let's get some energy going in here oh my god i'm so excited for this you guys are doing a fantastic event literally like logged off yesterday just pausing i had goosebumps up let's let's get you guys paid This was an experience that didn't feel like a webinar.
It's been a crazy day.
Okay, so that was Fable 5 .1. Let's take a look at Astra's version. This has been so energizing.
Yesterday, I did a workshop where I went all in on every little tool that I use.
Plot code is my primary driver. and then i built like my entire second brain slash personal agent on top of it i could go to bed come back it'll still be working you know 10 hours later to try to achieve that goal looks good it's great to be here you guys are doing a fantastic his lips aren't moving and super smooth you guys have been such an amazing audience today Now, this one, like I said, is really tough.
I think that they both did different things well. I think that Asher did a better job of putting like live clips in there. I think that Fable did a better job of kind of like gearing it up for the next event in October.
There were a few moments where there were some things that were off, like Fable had a few moments where the timing and the voices were felt like not very synced. Asher had one weird moment where it was showing a video of Russ, but it said Devin Kearns. And it was just like that was clearly off.
But I am going to give the slight edge here to Asher. But I will say I'm. Very, very impressed by both of those outputs.
I thought it was. I don't think it's the kind of thing that he would use to promote an upcoming event. I think he could do something much more polished.
And I actually think for these types of videos, you do want what we saw in the previous demo, which is a little more fine grained control of the tool afterwards. Having said that, there are times when I just need a quick video. In fact, sometimes even for YouTube or sometimes if I'm showing someone on my team what I've done with screen share and I need it edited down.
I just have now learned. give it to, frankly for me, it's now the Codex desktop app. I tell it what I want and it gives me a good enough video.
Nothing that I would yet use for promotion because as you saw, this wasn't great what he got back and going in and telling the AI to edit and change this and change that using text is too annoying. I want to be able to scrub through a timeline for that kind of change. All right, still good, still useful.
Let's go on to the next thing, which is analyzing meetings. So I want to go, I went from the like amazing, I can't believe we can build an app to like more day -to -day stuff that you will actually use day -to -day. Of meeting analysis.
So I asked it to look through all of my leadership meetings and my syncs with John and looking at other meetings that are relevant to the team. And I wanted it to understand, or I wanted it to tell me what are the biggest pain points, three biggest pain points, and one of the highest value automations that we could build to remove some constraints and help us scale faster.
So Fable gave me the biggest pain points of me being the single production and review engine, which is something that I've talked about 25 of 58 meetings. Certification delivery does not scale past one cohort. So we're trying to figure out what that will look like.
The critical constraint of the business is also a large cost center, 52 students, blah, blah, blah. And then we also have demand collapsed and every fix is blocked by unknown work and untrusted data. Weekly leads fell, blah, blah, blah.
And yeah. All of this makes sense as far as our three biggest pain points. And then the one automation it came up with was an AI customer success layer for the cert.
One live record per student built from circle activity, mastery check outputs and attendance. Students upload case readouts to an endpoint instead of DMing Pat. Weekly due dates and reminders to non -submitters go out through Gmail or Beehive.
I think this is something that I've wanted to do and that we've talked about a lot. And that makes a lot of sense based on all these conversations. And I also wanted to note here that Fable 5 .1 analyzed 58 meetings.
And if we go over to Astra, Astra analyzed 79 meetings. So it did get the context of 21 more meetings in here. But anyways, let's take a look.
The first one was that growth depends too heavily on a few acquisition channels. Yep, same pain point, I think, as Fable pointed out. The second one is that the customer journey requires too much personal intervention.
This appeared in an undefined AIS Plus onboarding call, a resurface during certification, and leadership still has to reconstruct status and clarify who owns what. Okay, interesting. So these are different pain points than Fable 5 .1 found.
Now, as far as the automation I would prioritize, it came up with the same thing. A customer success workflow, which would check each student's onboarding status, unanswered questions, available progress. identify missteps, prepare a follow -up, all makes sense.
Now, in this case, I think Astra won again. Before I even thought about the cost, which is absolutely absurd, Astra still won because I liked its recommendation about the leadership stuff and ownership. And I think that's something that you really want to make sure you have figured out within a team.
But anyways, if you look here, Astra was five and a half bucks and Fable 5 .1 here was 46 bucks, which just makes no sense to me. I do think if you haven't done this, connect your meeting note taker to your AI, ask questions, and you're going to get some real value out of it. Obviously, I think obviously you and I are not paying tokens.
So the price isn't as significant for us. But as somebody who is on whatever the max plan is on these tools. I do need something that conserves my tokens.
Just a few minutes before recording this, I was trying to create something for you and then I ran out of usage and I had to go and increase more. And so I do care about the cost. I do care about the time.
And interesting that I think that Astra has done so well here. Okay, writing. I got two more for you.
Writing is the next one. This is from somebody whose whole business is dependent on writing. It's a fantastic writer.
It one -shotted its own vibe check. I just I got to show you this like this. I should say, if you don't know the company, every run by Dan Shaper, who you see there, what they do is they create software using AI and then they write about the process.
They write about the tools and their business is selling a subscription to these articles like the vibe check, which is their initial impression of software that launched and from a subscription that includes software. So. This is a key part of their business, and he is really into writing.
This is the vibe check that it wrote for itself based on looking at our Slack and deciding what we thought. Vibe check. GBT6 Astra is a show horse.
OpenAI's new model has striking visual taste and a knack for consulting work. Its ambitions still outrun its judgment. It's a little harsh on itself, but this is good.
GPT -6 Astra is very good at getting carried away. Opening as new model out today builds beautiful things. I asked her for a 3D rendering of the Hadestown set and got a little theater with weathered green walls, hanging lamps, and a wooden stage ring with lights.
That's... That's very good. That's very hard to do.
I'm going to say this. For my taste, I don't like that kind of writing. I find that with Avery's articles, they go way too long with their sentences and way too dense with their information.
And I'm just trying to get a vibe check here. So I'd much rather it was punchy. Having said that, this vibe check written by Astra is very much in line with their style.
And I can understand why he's so excited about it.
I love it for writing. It is my constant companion. I feel like its sentences are crisp.
They're to the point. There's no AI -isms. There's no slop.
It's a very, very impressive writing model, and you need to check it out. And finally, let's take a look at a consulting deck. But here's basically what I asked, but here's basically what I asked.
And I should say, this is also done Fable and Astra. So you get a comparison and this is the kind of thing that you would do on a regular basis. So let's take a look.
I wanted a McKinsey level presentation on the state of SMB report for my Hercules consulting brand. So I needed this to be branded and I needed it to apply to our target audience with this fake consulting business. So that was Claude.
You can see that I gave Codex the exact same prompts. Now, immediately what I've noticed is that Codex asks more questions or GBT6 Astra specifically. asks more questions.
You can see right here, it already asked me questions, whereas Fable just took this and pretty much just ran. And that is something I noticed overall. So I just wanted to call that out to start.
So let's take a look at these two decks. Here is deck number one. I won't tell you which model it was yet, but right away, I love the way it feels.
It is capitalized. It has the branding. This feels pretty professional so far.
Now, I already did check both of these for the sources and the verification, and they are good from like... the fact perspective, right? The research they both did was good.
Now I'm looking more so at the way they actually formatted this deliverable. I think we're at a place now where GPT -6 Astra and Fable 5 .1, they're both so, so intelligent. You don't have to worry as much about fanning out agents and doing research.
It's more to me now about how do they come back and present this in a way that tells a story and it's something that I can actually like present over, right? Because that's what we want is we want to turn that messy data into something that's readable and something that's high quality. And right now what I'm feeling is that this is very, very branded.
It matches our color scheme. We have a footer down here, which I love. It's consistent.
We have like structure. There's columns. Like it did a good job creating this.
The thing that I think right now, though, is it's very, very wordy. As far as being able to present over something like this, it would be tough because there's so many words and it's tough to control like where people's eyes will be looking and what they'll be speaking about. So as far as a presentation deck, not the best, but as far as like a deliverable, you might turn in, you might send to someone and not present.
This is very good. And there's a lot of sources down here. This came out to 35 slides.
Now let's look at the other version, which right away, I don't like how this isn't capitalized. I don't know. I just feel like that feels very professional.
We have some images here and we have the color scheme as well, but let's flick through. We immediately noticed that this one is not as wordy. I immediately feel like this one's less professional though, because of the structure.
The other one had like, I don't know, you could tell in each slide, there was like columns. there was a little bit of like a, I don't know, there was a structure to it. Whereas this one feels a little bit more abstract.
It feels like one big element. You've got one big, you know, title thing. And I think that this one would be a much better deliverable if I was going to present in, you know, in a room full of people.
But overall, as we flick through, this one just feels less professional. The vibe I get is less McKenzie and it's more Google Slides or like Canva. I don't know.
There's also not like a consistent footer or our logo isn't on every slide. So, you know, this certified, it just doesn't feel very structured and as professional. Even the sources section is way smaller, although that does link the actual sources.
So anyways, this one was GBT6 Astra, and this one was Fable 5 .1. And I will say in this specific example, I liked Fable better. So what I'm going to do here is I'm going to give Fable the win.
So I'm just going to put this down here so we know Fable won this experiment as far as just the output. But let's look at the... time and cost fable was 37 minutes and astra was 23 minutes and fable was 26 bucks and astra was 12 bucks so astra was a little less than half the cost as well as you know 14 minutes faster Okay, so overall, here's what I have seen.
Both models are really good. When it comes to computer use, though, man, OpenAI has got the lead. And if you're not having it control your computer or your browser, I think you're really missing out on the power that they have.
And writing is, from what I'm seeing here, another win for Astra. Having said that, when it comes to design like this and some of the previous examples we've seen, I think both models need some kind of help. And so you should have some...
some tool that will help you guide them in the design style that you want alright if you found this useful I've got another video for you right here and I'll see you in that one
The Hook
The bait, then the rug-pull.
Eight different creators put OpenAI's GPT-6 Astra through its paces this week, so one host watched all of it, ran the model himself, and boiled it down to a single question: where does Astra actually beat Fable 5.1, and where is it just hype?
CTA Breakdown
How they asked for the click.
VERBAL ASK
07:39product
“You can go try it right now at zapier.com slash MCP.”
Mid-roll sponsor read for Zapier MCP, framed around AI-agent access control rather than a generic ad; the host also plugs his own next video at the very end.
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
Two hosts run through ten GitHub repos and skills that fix one narrow AI-coding annoyance each, then argue about whether skills are already becoming obsolete.
Corey Ganim and Andrew Warner speed-run a week of AI launches, an AI QVC host, an endless AI livestream, and a phone number for your agent, and call out OpenClaw 2.0's rough demo along the way.
Ben Senescu built OpenSEO to kill his own Semrush bill, wired it straight into Claude, and now watches open-source contributors fork it into CRMs and content tools he never planned to build.
00:16
00:22
01:03
01:40
02:06
03:01
03:21
03:41
04:22
04:43
05:04
05:25
05:46
06:28
06:48
07:08
07:28
08:10
08:31
09:15
09:32
09:51
10:44
11:06
11:27
11:49
12:33
12:55
13:16
13:39
14:01
14:23
14:45
15:02
15:48
16:11
16:44
17:03
17:29
18:07
18:33
18:59
19:42
20:02
20:22
20:42
21:02
21:22
21:42
22:08
22:41
22:52
23:20
23:39
24:04
24:43
24:54
25:17
25:37
26:14
26:35
26:43
27:33
27:45
27:57
22:57
16:34
27:32
19:02
27:14
21:02