that basically means that if you have something like VS code installed, then you open up like the Claude Code plugin that you should also have installed, something like this. Then you can drag any file, for example, the schema. json file into here, but you have to hold shift to drop it into here
Поиск по транскрипциям
soon. Hey, editing Eigor here. Just two really significant updates that happened since the recording. One is exactly what you just saw in Claude Co-work. They added to Claude Code 2. You could basically remote control it from your Telegram now, just like OpenClaw. Okay, that
models have different performance characteristics depending on the specific challenge. " Yeah, I would agree with that. Like for example, when it comes to UI, Claude is the best. coding stuff in one shot, it seems like Gemini 2. 5 Pro is usually the best. And then Quen, it might be like a state-of-the-art model...just check that out. So go inside Claude here. Run exactly the same prompt. It's a lot faster to respond as well, Claude. Claude is going on a rampage right now with the coding. Look at that. It's just blasting it out. What we'll do in the meantime is just test this out for something else
things. I'll put on screen now some prompting tips specific to the Claude 4 models that was shown in the code with Claude conference. By the way, in case you didn't know, Anthropic also does have a pretty good prompting guide as well as a prompt generator on their console. So, I'll also put the link...huge difference in the responses that you get. Now, before I move on to showing you the true power of the cloud for models, which is through building and coding through coding agents, I want to explicitly call out these cloud form models are definitely better. But if you're only going to be using it through cloud
bucks a month gets you the smart version. And there's a pro tier coming for the full power. But competition is heating up. Claude from Anthropic. It's beating GPT4 at coding right now. 72% on benchmarks versus GPT4's 54%. Google's Gemini 2 million token context window. That's twice what GPT5 promises. OpenAI's market share...switching between modes. One unified understanding. Upload a video, ask about specific moments, get perfect answers. Competition is fierce though. Anthropic doubled market share to 24%. Enterprises love Claude for coding. Google's pushing hard with Gemini integration everywhere. Even Apple's getting serious about AI. Open AI still has advantages. 500 million weekly chat GPT users. 92% of Fortune
GPT5. 2, it's performed worse than writing content with GPT40, it struggles to understand user intent as well as Claude or Gemini. He's failed all the coding tests versus Claude or Gemini. So, it came last. It couldn't write in a question mark on the content that we produced and it doesn't write headlines for articles...attrition include engineering etc. And this is GPT 5. 1 thinking versus GPT 5. 2. So this is formatting in spreadsheets and slides. Apparently it's better in coding although it seems to have got. And it says GPT 5. 2 represents the biggest leap for GPT models in a genic coding since GT5. And I just
simple and this is what I want to end this video on because look at that we had this little chat inside of Claud where it came up with the various codes and maybe this conversation went on even further and there's a whole lot there you're not going to be returning and reading that conversation ever...just straight up text that's what we'll do in this case so what I'll do here is I'll just copy my entire ation with Claud including the code how about that just say copy on this and here inside of notebook LM with a free account I'll just paste the copy text right here wonderful
remember it's available on all clawed plans including the free tier now let me share some real world feedback from companies that have been testing Claude 3. 7 Sonic cursor a coding platform said there's a reason Claude is the default model for all cursor users anthropics
want to make a few changes to the website locally on my own computer. Let's ask Claude to download the file and then open it up in VS Code. Claude clicks the save to file button, opens up VS Code, and then finds the file within my Downloads folder and opens it up. Perfect. Now that the file
Claude, just cut off access to its models for competitors. not users or random developers, but specifically its competitors. Claude stopped working inside thirdparty tools like Open Code and more importantly, Claude stopped working for XAI employees internally. And we know that for a fact because XAI co-founder Tony Wu literally told his team in Slack. He said, paraphrasing...Claude models aren't responding in cursor. Curser says that this is a new enthropic policy for major competitors. He even told the team, "This hurts productivity short-term, but it pushes us to build our own stuff. " Which is a very calm way of saying, "We just got cut off. " And I want to be very clear here. This
Claude 2. 0? It's an autonomous agentic coding platform powered by Claude Sonic 4. 5 designed to transform how developers build software. Now you don't need to be a developer to be able to use this. Right? So for example, if you want to use this and you have no coding experience, that is okay, my friend
This someone else recommended this. I'm going to say code in HTML. a PS5 control. Right now, the only good AI that I've actually seen code this properly is with Claude, right? So Claude, for example, if you give it the same prompt, it can actually do a pretty decent job of coding in HTML a PS5 controller
that's hyper precise and hypersp specific. The structured steps will return to Claude via very similar flow to what I showed you guys with the design. And then Claude will execute each using hyperspecific tools. Maybe if you're teaching somebody how to build something on Blender or Figma or something like that, you just give it access...this anti-gravity uh kind of container. Then inside of it, I use clawed code. And so in that way, I'm actually using a Google wrapper around a claude code or anthropic extension that's communicating with a claude or an anthropic model. If you guys want to replicate the setup, it's as simple as just opening
Grock 4. For coding on Claude, what mode do you use? So, I was using Sonet 4. 5 for that. That's the one I typically use. I think Opal is fast and it's better on the API cuz it's cheaper, but Claude 4. 5 Sonet is the most powerful one. Let's see what we got here...makes sense. Maverick says I don't at least fast or you can switch to instant I think on then it's better but still Claude for me personally there's one on the coding test and also for the poem test so far. So I'm going to go with yeah with claude for both of those. Okay
catch the coin. Display score for each C coin. Game ends after 30 seconds and shows total score. Provide full HTML plus JS code with minimal CSS for a functional game. Again, use Claude Opus 4. 5. It's the best for producing working playable code that actually runs. Here's what blows my mind about this. You can build...tell you why this matters more than you think. We're at a point where AI models are getting specialized. Grock for creativity and real-time info. Claude for long coding sessions, Gemini for hardcore reasoning, and platforms like client are catching on. They're giving you the ability to use the right tool for the right
Claude Code 2. 0 build anything. Today we're going to be looking at Claude 2. 0 and the new agentic release of this. This is using Claude Sonet 4. 5. And this is basically like a coding agent in your terminal. You can give it directions. So for example, if we open up the terminal here
Claude is Claude and Gemini 2. 5 Pro. You can use the latest models for free and code with them for free. So you can use Claude inside the chat or code for free and also Gemini astudio. google. com. With open router you are going to be paying, right? So this is a paid model which is that...interactive keyword calculator tool. You were supposed to code that, but actually didn't build out the SEO calculator, which is the whole point of the experiment. All right, so here's what I'm going to say. They all did a terrible job. I'm going to say that Claude probably came in first there. GPD 4. 1 came
with cloud code, you want to follow the spec, to-do, and code process. So after exploring and selecting a solution with claude, I like to ask it to create a spec with the requirements, tech stack, design guidelines, and up to three milestones to execute. And then I go even further by reviewing the spec, making sure...very much still in planning mode, just writing like documents. We haven't written any code yet. But again, the more you plan, the more likely Claude is to actually build the code without any errors. Especially for people who are like not super technical like myself, it's very important to the planning step first. So, let's come
Claude reasons with you, adjusts on the fly, and stays in lock step as you iterate. Code in writing or brainstorm in design. Use it as a code in buddy by asking Claude to review and run your script inside an artifact. Paste your Python invoice script. Say, "Review and run this and watch it execute on dummy data...hits a bug, say you see key error, total do claude flags the error, explains the cause, edits the code, and shows the success output. You don't have to say debug this. It just does the loop for you. Run, catch, fix, rerun. If the fix isn't ideal, nudge it to try an alternative approach, swap a library
what you can do for Claude. I think when you're vibe coding you are basically acting as a product manager for Claude. So you need to think like a product manager. What guidance or context would a new employee on your team need to succeed at this task? I think a lot of times we're too used...tell them what are the actual requirements and specifications and constraints that they need to understand. And I think that as we vibe code, that becomes our responsibility to feed that information into Claude to make sure that it has all of that same context and is set up to succeed. When I'm working on features with Claude