Claude Opus 5 is INSANE!

Florida Opus 5 is just dropped my friends a thoughtful and proactive model according to Anthropic at half the price of Fable 5 so it comes close to Frontier Intelligence at half the price of Fable 5 and you can see how it performs on the benchmarks today we're already running it through GoldieBench so we'll test it out and see how it performs across 50 different tasks that'll be coming to you soon but if we have a look at Opus 5 here this is Fable 5 versus Opus 5 versus Opus 4.0 versus GPT 5.6 Sol and what it really excels at is agentic terminal coding so you can see FrontierBench and it's absolutely crushing at 43.3% versus Fable 5 at 33.7% and GPT 5.6 Sol at 34.4% it's better at knowledge work better at problem solving better at agentic search so those are four things that it's better and outperforming all of the other models right now when it comes to multidisciplinary reasoning it's kind of mixed on this one so is excelling and it's a lot better at computer use versus these other options maybe something coming from or desktop soon on that I reckon related to voice if and if chat GPT is anything to go by then we have for example agentic coding and business workflows performing really well you can see here Fable 5 is still beating it at legal and health and then biology of all things Opus 5 is excelling at as well bear in mind like the difference with for example Fable 5 is that it's only including 50% of your chat of your Claude subscription whereas for example if you are using Claude Opus 5 you can use 100% of your subscription usage of Anthropic to get the most out of it so you get frontier level technology in a way where it is not limited on your current plan and so look how it performs versus other agents so the thing to note here is like if you're using the API then the cost per task is cheaper right so genetic computer use performance for example by effort level and you can see here the Opus is cheaper even when the effort level goes up so for example this is Fable 5 in orange and this is Opus 4.8 so Opus 4.8 is more expensive than Opus 5 even when you raise the reasoning levels and then Fable 5 is obviously way more expensive than both of those GPT 5.6 Sol when you raise the reason as well is still slightly higher and it's the same for example for business workflows by effort multidisciplinary reasoning and also agentic coding also on ARC AGI 3 in evaluation where AI models must solve novel problems Opus 5's score is three times as high as the next best model so we can see here how well it's performing Opus 4.8 all the way down here so I mean the main thing to note here is like you're basically getting Fable 5 level intelligence according to this you always want to test it out yourself you're basically getting Fable 5 level intelligence on a model that is cheaper or you can use with your full subscription and if you look at the gap between Opus 4.8 and Opus 5 it's absolutely huge on most of the benchmarks here they've also said that Opus 5 is stronger than Opus 4.8 outside of security tasks but remains substantially behind Miphos 5 at developing exploits so safeguards are designed to allow developers to identify and fix software vulnerabilities whilst blocking high risk uses and it's available today so for example if you go over to Claude.ai it's available right there and then also we've already plugged it into our agent operating system so you can see here that we've got Claude Opus 5 running directly and then everything that we build plugs into the workspace plus it already goes straight into our memory system so that we can get the most out of this now bear in mind this just dropped like literally 15 minutes ago as of recording this so it's straight hot off the press i haven't had much time to test it yet but let's have a look through the announcement here and see what we've got bear in mind that really this there is a lot of pressure on Anthropic right now because number one they've messed people around a lot with Fable 5 then you've got GPT 5.6 which is absolutely crushing it and a lot of people saying they prefer GPT 5.6 to anything else out right there Grok 4.5 is getting better all the time and then you've got two models Kuen 3.8 Max from China which is right up there at frontier level with something like Fable 5 we've tested out on Goldie Bench and you also have Chinese models like Kimi K3 open source fully open source but super super powerful like right up there with Fable 5 as well so they have to drop something that's more affordable and more available just like Fable 5 and that's why Opus 5 has to come out also there were rumors of it being dropped yesterday but i think that was delayed also if you want to indicate what is it good for so apparently software engineering tasks pretty good good on frontier bench surpassing all other models on cursor bench 3.2 max effort the model performs within 0.5 Fable 5's peak score but at half the cost per task so it's like you basically get the same level of technology but half the cost and apparently it also achieves greater performance at a given cost than all other models on high x high and max effort some interesting ones as well so is i think that there's going to be some big upgrades inside cloud desktop and cloud co-work because if you look at what they're targeting it it's like they're seeing similar results on knowledge work and problem solving tasks here right so for example ARC AGI 3 an evaluation where the model has to solve novel problems Opus 5 score is three times as high as the next best model which is pretty awesome a Zapier automation bench which measures whether my tools can complete business tasks from start to finish Opus 5's pass rate is around 1.5 x the next best model for the same cost per task even at its lowest effort setting so Opus 5 passes more tasks than any other model and on os world which is a computer use benchmark so basically how well the agent can use your desktop your computer i've been testing out chat cheap voice for example earlier today Opus 5 outperforms every other model at given cost any given cost surpassing fable 5's best result at just over a third of the cost so it's just way way cheaper so it's the best and most cost efficient model on many of the related evaluations around again test it for yourself don't believe me don't believe the benchmarks test this stuff out again we are testing it out we've got loads of different builds going out for goldie bench right now we've actually got six different demos i can show you right now these are like one shot demos so if they're not very good that would explain why but let's see what it's like so this is the dragon realm game and actually that is super nice that is i'm going to show you some comparisons in a second but this is looking really really good so it's kind of like a skyrim style rpg game and like the graphics and everything else pretty cool right looks really nice also something that i'm noticing is like the gameplay is really interesting here so if we look at the way it's like wobbling around side by side as you go around i've never seen that from an ai model before so this is one shot like it's had one attempt to get this right and it's actually absolutely crushed it which is amazing and you can see the enemies you can see the gameplay i don't see any bugs the graphics are nice the details and i see ambience the vibe the feel of it is pretty cool as well even like the flashes on the screen and everything else so it's looking really cool and then when i walk around as well um looks really really nice i even like the background detail here with the northern lights super nice right the the snow drop and everything else i'll show you how that compares to for example fable 5 here and by the way if you want to like test any of these demos they're on goldie bench for every single model that i've tested out so you can just have a look for yourself and see what you think let's have a look at it so this is fable 5's version of dragon realm and you see it's just not quite as interesting not quite as nice so i mean look at this like the comparison is is massive the difference is massive the quality here is way better from opus 5 i'm not going to get carried away i still think you need to relentlessly test this stuff which is why we'll be running 50 different builds today um but is it looking promising is it looking good absolutely it's looking great so far so look so we've got a doom ray maze here this is this is nice this is nice um it it doesn't feel laggy at all the controls are nice as well let's go around the maze and see what we got here the gameplay as well is pretty fun as you saw when the the zombies popped up or whatever so here we've got another one yes i mean it's looking good like bear on it literally built this one shot just got it perfect one shot which is very rare so for example when we were testing out uh quen 3.8 max that was not a one-shot take that was not a one-take drake like that was you know a few back and forths orchestrating the agents getting them to do it right but this is looking good uh so that is two different tests that look really good here i want to be 100 transparent with you though don't want to bs you so if we have a look at this one i just tested out now not good at all it doesn't get past the screen so it's not like it gets it perfect first time around i just don't want you to be disappointed if you try it yourself um and i don't want to pretend like everything is amazing when it's not there's gonna be certain things that are just not that great so you can see the scores here and it's already being scored in terms of how it performs um now we've got like this neon snake game as you can see here pretty fun to use this is literally like a 3d snake game um kind of crazy like the the gameplay and everything else really really interesting let's have a look at this one this is like a sort of air combat game right a little bit hard to control let's try it again no it's nice it is nice like the there's something about it like the game i think it's the way it wobbles around when you you're flying and everything else i mean that's making me feel a bit dizzy to be honest but let's refresh that there we go all right yeah let's try the next one oh this is nice i mean actually one thing i'm seeing here is like it's struggling a little bit with the physics of the game it's a little bit hard to control but actually that's good the quality is good the vibe is good the colors are good the detail a little bit tiny little bit off um but this is some of the most interesting stuff that i've seen that ai create right like if you check out the other examples on goldie bench this is right up there with fable five undeniably um the gameplay the 3d graphics the fact that it just took this in one shot as well very impressive so i'm excited to see what else it builds out let's have a quick refresh see if we've created any others so far now that is six so far that we've built feel free to check them out yourself i think the dragon realm is looking the best so far really cool really cool idea um yeah can't wait to try out more to be honest so what we've actually done here as well is inside the agent operating system we have plugged in opus 5 and that was super easy so if you want to get our system with the full setup comment uh sorry go inside the air profitable link in the comments description or go to the air profitable to get that and that basically works with all your agents you got hermes you got anti-gravity you got claude all alongside here your full workspace everything built out as well which is pretty cool and then we've got the memory system and all these tools can be operated so for example like uh open design open montage or video editor etc these can all be operated by claude opus 5 and we've done all the hard work set up so you can just get the full system from us let's see what else we got here so they also say claude opus 5 is much stronger at verifying its work and iterating carefully until it succeeds in evaluations and early access testing we and users found many examples of frontier bench absolutely crushing it okay so basically what it's good at here is like loops where you don't prompt it you just give it loops and it loops around until it's done but so far pretty good there's a bunch of like security things here as well which i'm not too fussed about uh but here we've got for example the details so it's priced at five dollars per million input tokens and twenty five dollars per million output tokens the same as opus 4.8 so it's not increased in cost despite getting better um you know improvements it's also offered in fast mode which is good because i don't think you get that inside fable 5 so it can run 2.5 times the default speed which is great for building it just build a lot faster for you which is awesome as well however it's available at twice opus 5's base price so you get 2.5 times the speed but it's pretty expensive pretty expensive right there and there's two updates here as well so mid conversation tool changes on claude platform so you can now change the tools it can use within like conversation i don't think that's a big thing and then automatic fallbacks on the api so if you're using the api some things ah this is ah this is not good so i've seen this before with fable 5 where basically automatically routes to another model when it doesn't like what you're doing and so basically it automatically falls back and i think a lot of people will complain about that to be honest but again is it good yes have you seen the outputs yes is it inside the agent os yes you want to get it right now of course and um i think it's going to be an impressive model i expect there's going to be some people who are quite cynical about this model as always is i'm interested to see what people saying about it let's just have a cheeky gander see what people say on x about it whether anyone's complaining yet not yet but i think that's coming my friends let's have a look here arena ai say claude opus has landed in the arena okay so people are going to be testing it right now no complaints yet but i think they'll be coming so thanks so much for watching if you want to get our full system with the agent os claude 5 plugged in full workspace where you can save everything and come back to it later you got the memory system so that you can give instant context and stop wasting your time re-prompting claude and ethanels also you have all your agents working side by side which is awesome if you can't afford claude or if you don't want a subscription we also have free claude code and we have gpt 5.6 claude code so basically you can plug claude code into glm 5.2 kimi k3 quem 3.8 or gpt 5.6 so you get the harness as well if you prefer to use that and then you see all these amazing tools like open montage video editor open design the seo workflows the video agent everything else you can see right here we've got inside the agent operating system so feel free to get that it's available inside our community the ai profit in the comments description or go to the air profitable.com and inside this community we just focus on helping you learn save time and grow with ai automation so you can post your questions inside the community we answer them uh date today with video tutorial inside the classroom you can get access to all of our best trainings including our new daily updates we'll be dropping a claude 5 opus masterclass inside here as well to help you as much as we can and then inside the classroom get all of our best trainings including courses step by step you know you've got loads of people that are just online 24 7 to help you plus i'm ready to help you as well you can direct message me with anything you need to and then you can get our agent os system here which is about to update to with opus 5 so you can get the video tutorial up um and you can get the zip file to install it plus all of our best guides so you can see inside the calendar you can jump all we can close your calls get help support time and share your screen meet other cool people using claude and opus and other tools and then inside the cat inside the map you can actually meet people in your local area who are building with ai agents like you so see you inside there cheers for watching see you in the air profit boarding