Okay, so Claude Opus 5 literally just dropped. Now, obviously I'm going to spend the rest of my day playing around with it, comparing it to Fable 5, and I'm going to bring you guys an experiment. So, stay tuned for that later today. But, I wanted to just real quick come in here and just show you guys this cuz I think it's really, really interesting. On several coding and knowledge work evaluations, Opus 5 is the new state of the art. So, look at this first benchmark right here. Agentic terminal coding. Opus 5 has 43%. Fable 5 had 33%, and Opus 4.8 had 21%. So, this is a major jump in agentic terminal coding above Opus 4.8, but also above Fable 5. Same thing with knowledge work. This thing crushes Opus 4.8. I don't exactly know what the novel problem-solving benchmark is solving, but look at this. It went from Opus 4.8 at 1.5 to 30% at Opus 5. So, anyways, these benchmarks are insane. You can also see here computer use, which I always would default to Codex. I always thought that GPT 5.6 or 5.5 with Codex was just the absolute best with computer use, but look at this with Opus. Opus 5, I'm going to have to try it out with computer use. Anyways, what's really interesting to me is this efficiency because as you guys know, Opus is half the price of Fable, and Fable will burn through your credits like nobody's business. But, look at this. Agentic computer use. Opus 5 performs better than Fable and for cheaper. Look at this. Agentic business workflows. Opus 5 is better than Fable and for cheaper, which is just really, really interesting to me. Multidisciplinary reasoning by effort level. Opus 5 does better than Fable and is cheaper than Fable. So, like I said, guys, I'm about to run a ton of usage credits. If you look at this right here in my Claude, I am already at my weekly Fable limit, and I'm about to go into, you know, thousands of dollars in extra usage credits, but I'm going to run a ton of experiments here, and I'm going to bring you guys my consensus. But, I just wanted to come in here and just like show you guys that real quick. It's really interesting. So, if you real quick go into your VS Code, go into Claude Code, update it, get on Opus 5, and just start running your skills and start running things and seeing how it feels. Because as we know, these benchmarks are fun to look at, but always take them with a grain of salt. It always matters on your use case, the way you talk to agents, the type of work you're doing, things like that. So, Claude Opus 5 provides greatly improved performance for the same cost as its predecessor Opus 4.8. Opus 5 excels at valuable software engineering tasks like the Frontier bench, on Cursor bench. I really want to see how it performs on the Deep suite because that's the one that's been like a pretty true like leading indicator. But, the whole cost thing is just really interesting to me compared to Fable 5. Cuz I know a lot of you guys were kind of like, "Oh man, it it sucks that we're going to be losing Fable or that, you know, we only get a certain different limit for it, but it's overkill for most things and it's really expensive. So, if Opus 5 is better for like general knowledge work, then that's a huge win. And this is really exciting to me. Opus 5 is much stronger at verifying its work and iterating carefully until it succeeds. I heard this really funny analogy on X, which is honestly pretty true. They basically said like, "Hey, Fable 5 is like a wise old owl. It's really good at planning and managing and ideating and coming up with ideas, whereas GPT-5.6-Soul is like a Rottweiler because that thing will grab onto a task and it will shake it and it won't let go until the task is done, which I definitely felt. It was spinning up more tests and it was better at verification than Fable was, right? But now, maybe Opus 5 saw that and was like, "Oh wow, we need to work in some of this GPT-5.6-Soul verification." And now, if we have that, that's huge because verification is kind of what powers all of these agentic loops that everyone's talking about. And if you're using a model that's really, really good at verification, your loops are going to be better and your outputs are going to be better. So, I'm really excited to test out Opus 5 with some verification. And here are some examples where they gave Opus 5 some real scenarios and had it find bugs, verify, take different approaches, and ultimately give a good output. But, anyways, this is now completely available to everybody on all platforms, same exact price as Opus 4.8. So, like I said, I'm about to jump in right now and start testing the heck out of it against Fable 5 and I'll bring you guys that full experiment later today. So, keep your eyes peeled for that video. But, anyways, I'll see you guys there. Thanks, everyone.

If I ran an accounting firm, here's exactly what I would do to automate my entire business. I'd first get my partners off of work that doesn't need a partner. Because right now, the most expensive people in your firm are spending the first 5 days of every single month just rebuilding the same client report. Watch this. So, the second the books close, an agent pulls the numbers straight from QuickBooks, it drops them into your reporting template, and it writes the first draft of the commentary. So, what moved this month, what's off versus last month, and what the client needs to be told. So, every number is pulled from the source, never just invented and every line it links back to where it came from on an audit trail. It then lands into a review queue where your partner edits the story and sends it off. Now, this always drafts and never sends anything itself. So, nothing reaches a client until a human approves it. If you want this set up for your firm, just comment need and I'll send you more information.

All right, so Claude Opus 5 is here and if you start to look at the benchmarks, it's really interesting because it shows us that for a lot of things that I care about, it's actually better than Fable and it is half the cost of Fable. Ultimately, Fable 5 is still Anthropic's most impressive and, you know, strongest model. But for a lot of these things, you know, I've realized when I'm doing knowledge work and when I'm building, you know, my videos or my research or whatever it is, Opus is more than enough power than what I need. And when you look at some of these charts, it's really interesting because it shows on things like the Frontier Bench and the Cursor Bench and this coding agent index that Opus is actually outperforming Fable and it's cheaper. And this really shocked me. So, obviously, I like to take all this stuff with a grain of salt. It's fun to look at and it's good to look at, but you want to actually get your hands dirty and run these models through your own actual workflows. So in today's video, I'm just going to break down a bunch of different experiments that I ran with Opus 5 versus Fable 5 and break down things like the cost, the time, and the tokens so that you can start to understand where you should work in these different models within your workflows. All right, so pretty much all of the experiments that I've been running today that I'm going to show you guys, I did within Claude Code, which means we're comparing the models, but also inside of the Cloud Code harness. and the variable is the same so it doesn't really change too much but I did do a few tests where I was actually in clawed chat and I was just you know seeing how they felt without a harness wrapped around and let me just show you one quick example so here I asked Fable 5 and Opus 5 to generate me an Excal diagram that accurately and visually explains how semantic search with vectorization works on a large data set for AI agents and it's interesting here because there's no skills that it can use and it doesn't have any context of me or you know any really way to verify all it did was it spit out a um JSON file of Excal for me and then I pasted it into Excal. So, here's what we got. Fable came back with this version over here where we see we've got like our indexing pipeline. We have a large database and it looks like it actually misspelled this right here, which is interesting. Large. Oh, data set. Okay, it was just like not expanded enough. Same thing over here. Vectorize. And this is part of the whole um it had no way to verify. And as you guys know, if you've been kind of building agent loops and stuff, verification is so so important. So anyways, large data set, we chunk it up. We vectorize it with an embedding model. We then get our embeddings, which is just like the numerical representation of the data. We put it into a vector database here. And then we can actually start to search. So we've got similarity and that's on, you know, points being close together. So your question lands here and it would grab the k nearest neighbors and different meaning is farther apart. We've got different clusters here. And then we come down here to the actual query. So if the user asks how I get my money back, the agent searches the knowledge base and then it does semantic search. It looks it up with the actual query vectors and we get the matches back. So pretty accurate. I will say though, Opus' layout seems a bit more organized, right? Like it's it's got boxes and it's got I mean this might not be as visual. You could argue you could argue that Fables was more visual, which you know I think that that's true, but this definitely feels more organized. it feels a little bit more detailed as well. So, that's just a very subjective exam. A lot of the stuff that I'm going to be talking about today is just really opinionated and subjective, but I'm going to still give you my own thoughts. So, in this example, I think that if I wanted to teach someone, I probably would take Opus 5's version here. Okay. So, let's start off with the first test I ran, which was basically giving them a huge codebase and having them look through any bugs and looking through like the expected behavior and some instructions like that. So, it had to do some exploration here and help us out, right? So, I set the goal and I gave it this prompt. And then what I did is I had Codeex review the output that Fable 5 gave us and that Opus gave us. So, real quick before we look at the results, Fable took about 11 minutes and it costed 5 bucks. 30, whereas Opus here took 13 minutes, so a little bit longer, but it was cheaper at $4.22. You can also see the breakdown here of input and output tokens and like what models they use and stuff like that. But, let me switch over to Codeex here. This is the actual result. So head-to-head, they both pretty much passed everything, which is great. But Codex thinks that Fable wins here because Fable's production patch is exactly the onelin