GPT-5.6 is HERE (WOAH)

Here's the thing about GPT 5.6. It is

truly a massive leap from GPT 5.5. It

does not seem like they should have just done a dot upgrade. They have

effectively squeezed every drop of juice out of the GPT 5 training run that they possibly could and the result is one of

the most effective models and capable models that I have ever used. One of the first loops that I gave it was super simple. Look at this. /goal make an

simple. Look at this. /goal make an Excel clone continue until feature parity. Here is the thing about 5.6 plus

parity. Here is the thing about 5.6 plus Codex. It has all the tools in its tool

Codex. It has all the tools in its tool belt to do this really well. Just a

simple prompt which is effectively eight words long ran for over five days before I manually stopped it. And here is what it created. This is the Excel clone and

it created. This is the Excel clone and it has a lot of the features that Excel has. And if I would have let Codex keep

has. And if I would have let Codex keep churning through tokens it probably would have gotten even further. All

right. So here very simple we have a list of numbers. You can easily sort it by ascending descending. If I add another number in

descending. If I add another number in there, same thing. And everything is just super fast, works really well. We

have all the formulas. So let's say equals this plus this nine. And if you double click into it, you get the formula up in this input bar up here

like normal. We can add another one

like normal. We can add another one minus 43. Hit enter and everything just

minus 43. Hit enter and everything just works. So we have a bunch of standard

works. So we have a bunch of standard Excel features. We have data validation

Excel features. We have data validation conditional formatting. We have tables,

conditional formatting. We have tables, find and replace, obviously toggling and sorting very easy. We have all of these different formula that you can use right

from this single page HTML app.

Remember, Excel is a massive piece of software and a lot of the functionality has been condensed down into this very simple app. You can go try this out,

simple app. You can go try this out, which I'll drop a link down below for that. Shout out to hear.now for hosting

that. Shout out to hear.now for hosting all of the demos you're going to see today. It also has really deep analysis

today. It also has really deep analysis capabilities, including pivot tables.

Look at how easy this is. Simply select

some data with headers. You get all the rows sorted out. You get filters. You

can create the pivot table. And look at that. We actually have a full pivot

that. We actually have a full pivot table. Done. And again, this ran for 5

table. Done. And again, this ran for 5 days. Imagine how long it took to

days. Imagine how long it took to actually create Excel. Took years and years and years to get it to where it is. And obviously this is a subset of

is. And obviously this is a subset of the total functionality. And obviously

as you can see here, there are some rough edges. But again, I stopped it

rough edges. But again, I stopped it after 5 days. And it was not anywhere close to being done. And here's the interesting bit. It used computer use.

interesting bit. It used computer use.

It opened up Excel on my desktop and would just go back and forth between doing something in Excel, the actual Excel, and then recreating it in this

new cloned version. And that is one of the things that GPT 5.6 is so incredible at. Browser use and computer use.

at. Browser use and computer use.

It is phenomenal at browser use. I

really cannot overstate that. Codex's

browser has increasingly become my default, my main driver browser because I can just get stuff done with it. I can

open up Gmail and have it sort through my emails. I've done complex DNS record

my emails. I've done complex DNS record changes easily with just one prompt with Codex browser use. And of course I had it clone Minecraft. I did {slash} goal

once again and I just said, "Create a clone of Minecraft, feature parity." And

it went for something like 7 days before I finally stopped it. And it took only about 1 day to get what looked like actual Minecraft. But what it kept doing

actual Minecraft. But what it kept doing was going deep and would build out parts of the world that didn't exist before.

Would build out mobs that came straight from the actual game of Minecraft, different biomes, and it would just continue churning and just trying to get to feature parity. And it was just so

impressive. And watch this. This is

impressive. And watch this. This is

definitely the best version of Minecraft that I've ever created using AI. All

right, so here we go. You can see the shadowing is really nice. Here I can mine the grass blocks pretty easily.

Really cool 3D animation when you do that. You can pick up all the blocks

that. You can pick up all the blocks that you just mined. Here we have miniature carrots. There's some carrots.

miniature carrots. There's some carrots.

We have farmland. And yeah, and it just works really, really well. Here's a

tabby cat. Here I broke the glass. And

so it is just incredibly impressive. It

is a very full world, more full, more realistic than any other Minecraft that I've ever created. You can generate different seeds. Here it is. Here's that

different seeds. Here it is. Here's that

new world I just created. Here is my full inventory. Very easy to use. Very,

full inventory. Very easy to use. Very,

very cool. And of course our friends and partner on this video, Box, has put together their own benchmarks on enterprise-grade work. Let's take a

enterprise-grade work. Let's take a look. We have Box AI complex work eval

look. We have Box AI complex work eval for GPT 5.6, Soul, Terra, and Luna. And

Box's benchmark tests real knowledge work, like reading documents, reconciling numbers, doing due diligence, and reviewing expert output

for errors. We have GPT 5.5 on the

for errors. We have GPT 5.5 on the accuracy. 63.3, Terra at 59%. So

accuracy. 63.3, Terra at 59%. So

obviously a drop from Soul, even a drop from 5.5. We have 5.6 Luna basically

from 5.5. We have 5.6 Luna basically getting the same score as Terra, but much faster and much less expensive.

Here we have industry subsets. We have

the public sector, life sciences, and healthcare. All three Soul dominated GPT

healthcare. All three Soul dominated GPT 5.5. Thanks to our partners at Box for

5.5. Thanks to our partners at Box for putting together this awesome benchmark.

I'm going to drop a link down below where you can read more about their benchmark and specifically how GPT 5.6 did on it. All three of these models are coming soon to Box AI, so go check them

out. I'll drop a link down below. The

out. I'll drop a link down below. The

pricing is also much better for GPT 5.6.

Not only is it less expensive, but it uses less tokens to get to the same result. So, $5 per million input tokens

result. So, $5 per million input tokens versus $10 for Fable, much cheaper on cash hits, and 30 versus 50 dollars per million output tokens. So, again, you're

just paying less, and it just feels like 5.6 has a more direct line of sight to accomplishing the task versus Fable.

Now, with all of that said, Fable is something else. It feels much more like

something else. It feels much more like a brand new model, something that I hadn't used ever before, and it is very impressive. It sees around corners

impressive. It sees around corners better than GPT 5.6. Here's the best way to explain it. 5.6 feels like the absolute pinnacle of an existing model,

where Fable feels like we're just scratching the surface on what's possible on a brand new training run.

That's the difference. Here's an analogy for you. It's kind of like GPT 5.6 is

for you. It's kind of like GPT 5.6 is the most souped-up Honda Civic you've ever seen. Every single horsepower has

ever seen. Every single horsepower has been squeezed out of it. The tires are optimized for speed. There's a spoiler, everything. And Fable is like a Ferrari

everything. And Fable is like a Ferrari that hasn't been touched yet, fresh [snorts] off the manufacturing line, unoptimized, and the potential is just

so much higher there. So, I put together a full review. I have a bunch of demos of things that I've built, including an actual operating system. We have a Rube Goldberg lab that you can use. All of

this is available on it. I'm going to drop a link to the full review down below. And here's the other thing that

below. And here's the other thing that is very different about GPT 5.6 versus Fable. It comes in three different model

Fable. It comes in three different model sizes, Luna the smallest, Terra the medium, and Sol the largest. But even

within those, you have multiple levels of reasoning. And if we choose Sol, you

of reasoning. And if we choose Sol, you get all the way up to Ultra, which is, you know, basically a quota burner. And

so we've been talking a lot about model routing lately, where maybe you're using Fable for planning, and then you can actually call Codex from within Claude Code and delegate off to GPT-5.5. And

now you can kind of just do everything with the GPT series of models. You plan

with Sol, maybe you do most of the implementation with Terra on, you know, high reasoning, and then for stuff like deploying or other kind of low requirement work, you offload to Luna.

And so now you have all these different sizes, all these different thinking effort settings for these different models, and you can come up with a really nice skill to delegate between

them. And by the way, I wrote a skill

them. And by the way, I wrote a skill just for that. And if you want it, I'm going to drop a GitHub link down below, so you can delegate all within Codex, and save yourself a bunch of your quota,

and get basically the same quality performance.