Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7 💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see! 📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1 🔥 Become a Patron (Private Discord): patreon.com/WorldofAi 🧠 Follow me on Twitter: twitter.com/intheworldofai 👾 Join the World of AI Discord! : discord.gg/NPf8FCn4cD Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
I'm not even a dev but I have been waiting for the day when I can change from Claude 3.5 Sonnet to something else within Cline + VS, I guess the time has come.. Please let me know here how you went @CharlotteLopez, would be very much appreciated 👍
Deepseek appears to have a more limited context window compared to other models, causing it to become ineffective with Cline during extended conversations.
Hey, nice video! I wanted to try it out, but where do I have to paste the hyperbolic API (for those free credits)? Or, when simply entering the deepsek API in Cline / VSCode, and follow your steps in the video, I get the info "402 Insufficient Balance". What am I doing wrong? Thanks in advance 😊
@habibi750 with VS Code and Cline. What I've learned since that comment I made. Is give a very high overview descriptive prompt of what you want. Use claude for the initial build out. And any individual prompts that you want after that use deepseek.
It's best to use it through open router. You don't have to worry about running into the context limitations when using it through deepseek. I've been going strong in one chat stream with no issues for a while now.
Not these ... Make a big size complex full stack application with react frontend, any framework for backend along with jwt authentication and postgres as database. Is that possible?
Easily, im building a whole Ai agent framework with multi agent collaboration sysytem. You just need to understand what you are doing. Ai is elan enhancer, not a replacer (yet)
While this demo impressively showcases DeepSeek-V3 and Cline's capabilities for automated code generation, labeling it as "full-stack" development feels misleading. True full-stack applications require much more than UI components and standalone scripts - they need robust backend systems with database management, API endpoints, authentication, security implementations, and proper deployment infrastructure. The video beautifully demonstrates frontend automation and basic scripting, which is valuable in itself, but let's be precise with our terminology. A more accurate title would focus on the actual strengths shown: automated frontend development and script generation. Still, the cost-effectiveness and seamless integration between DeepSeek-V3 and Cline are noteworthy achievements for the no-code movement.
@@wisecat324 I'm actually a human viewer sharing genuine technical concerns about full-stack development claims. My comment comes from real development experience - proper full-stack involves complex backend systems, databases, authentication, and deployment infrastructure that weren't shown in the video. The examples were impressive for frontend/scripting automation, but calling it "full-stack" oversimplifies what full-stack development entails. Not every technical critique comes from AI - developers often discuss these nuances to maintain accurate technical terminology in our community. Happy to have a constructive discussion about what constitutes true full-stack development!
@@wawaldekidsfun4850 Yeah I agree with you on that. And I didn’t think your response might be AI generated because of your critique, but rather, I thought so due to the phrasing, some description words that I come across frequently when chatting with AI like “noteworthy”, and the perfect punctuation, which reminded me how AI generated text is usually is. But either way, you have valid points.
I love your content. Thanks for showcasing this technology so quickly. Could I make a request that you zoom in on your screenshares? I have a big new phone and I still have to put my eyeballs directly on my screen to even have a chance of reading the text. You leave a lot of wasted room on the sides, when you could be zoomed in instead. I appreciate the chart at the beginning of the video. I could actually read it this time. In the last deepseek video it was too small for me. Thanks for all you do. No hate. Just a request. To be fair, it probably looks fine on desktop, but I might not ever watch it there.
I will have to try it out so far some of the other ones haven't been to well with coding trading type bots and me not knowing much about coding doesn't help lol. Sometimes when your code gets a little more elaborate or multiple features they start to fail.
Can it read your whole file/app structures and/or edit specific files that associated? like windsurf that can read the whole codebase before implement code.
It sends your file structure to the selected LLM. Then while solving an issue or generating the code it keeps reading the files based on the file names. It does not have the capability to "search" your codebase as windsurf does
try this: 'can you analyze my whole site', 'create a readme file that shows what my site is and what it needs', 'in the readme file give me some suggestions on how to improve my site', 'in my readme file create a section of what my site needs to get done, create a checkbox, leave unchecked what needs to be done and check the box if it has already been done, everytime something gets done, add a check to the checkbox'. By doing this, everytime you use the AI with a brand new context, just ask it to review the readme file. It's like giving the ai memory.
I realized one thing, knowing the code is half the battle, you also need to know the topics for which you write this code. I tested this Chinese product for 24 hours and came to complete disappointment, you need to conduct tests of the code + topic and compare them. Then the analysis will be correct. It is an order of magnitude weaker than Cloud, which works through windserf
I've found it to be lacking compared to Claude. Yes ita cheap and yes cline can auto run but 30 min of python tests and it was still failing. Moved back to claude/winsurf and started over... Took 3 minutes to pass the tests and move on.
[Must Watch]: Groq AppGen: FASTEST and FREE Way To Create Full-Stack Applications! (Opensource): ruclips.net/video/QvnZ_4pz8Uc/видео.html AnythingLLM: All-In-One AI Platform! Fully LOCAL Chat With Docs + Do Anything with Local Agents!: ruclips.net/video/WsQLC1jOO1U/видео.html CodeGate + Continue: This FREE Extension BEATS CURSOR! (Opensource + Local: ruclips.net/video/mcEcJ_QgYLw/видео.html
Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7
💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see!
📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1
🔥 Become a Patron (Private Discord): patreon.com/WorldofAi
🧠 Follow me on Twitter: twitter.com/intheworldofai
👾 Join the World of AI Discord! : discord.gg/NPf8FCn4cD
Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
@intheworldofai Am beginners for AI but I'm enthusiastic about it, would you recommend me from where to start and what to learn please?
Deepseek-V3’s speed + Cline’s simplicity = Developer’s dream. Can't wait to see this real-time magic. Great tutorial!
When you go deeper, you see disappointment, if you compare it, for example, with winsurf.
I'm not even a dev but I have been waiting for the day when I can change from Claude 3.5 Sonnet to something else within Cline + VS, I guess the time has come..
Please let me know here how you went @CharlotteLopez, would be very much appreciated 👍
There are only problems with the context length
Deepseek appears to have a more limited context window compared to other models, causing it to become ineffective with Cline during extended conversations.
🧐😠🤨
So true
What is Cline?
@@johbyuhan vscode extension that make you like a programers
yes only 64k it is very less..
I can not wait to build! 🎉Thanks for sharing❤
Hey, nice video! I wanted to try it out, but where do I have to paste the hyperbolic API (for those free credits)? Or, when simply entering the deepsek API in Cline / VSCode, and follow your steps in the video, I get the info "402 Insufficient Balance". What am I doing wrong? Thanks in advance 😊
me too
As good as it seems, context is a let down.
I'm a total noob, can you explain what you mean?
I've used it today. Its okay. Context window is limited which is a killer if you're doing full stack. Claude logic is still superior.
How are you using Claude?
@habibi750 with VS Code and Cline.
What I've learned since that comment I made. Is give a very high overview descriptive prompt of what you want. Use claude for the initial build out. And any individual prompts that you want after that use deepseek.
for me its not working it is saying : 402 insufficient balance
same
me too
It's best to use it through open router. You don't have to worry about running into the context limitations when using it through deepseek. I've been going strong in one chat stream with no issues for a while now.
Non yt product
Cheers ! Thanks for sharing !
Thank God! Using Claude 3.5 in cline feels like having another car note.
Thanks for the video! really appreciate it
I heard DeepSeek-chat has only 64k context window which is not enough often, but that there`s a version with 128k, is it DeepSeek-coder?
Can we adjust the token? It limited the token and cannot complete the work yet.
Not these ...
Make a big size complex full stack application with react frontend, any framework for backend along with jwt authentication and postgres as database.
Is that possible?
Easily, im building a whole Ai agent framework with multi agent collaboration sysytem.
You just need to understand what you are doing. Ai is elan enhancer, not a replacer (yet)
@@roynijland7386 which ai tool you're using to build full stack applications?
Cant see /pass the payment method to make a top up... how come?
While this demo impressively showcases DeepSeek-V3 and Cline's capabilities for automated code generation, labeling it as "full-stack" development feels misleading. True full-stack applications require much more than UI components and standalone scripts - they need robust backend systems with database management, API endpoints, authentication, security implementations, and proper deployment infrastructure. The video beautifully demonstrates frontend automation and basic scripting, which is valuable in itself, but let's be precise with our terminology. A more accurate title would focus on the actual strengths shown: automated frontend development and script generation. Still, the cost-effectiveness and seamless integration between DeepSeek-V3 and Cline are noteworthy achievements for the no-code movement.
Hey ChatGPT, didnt know you can watch videos and comment on them as well. Now you have opinions too? What a time to live in
@@wisecat324 I'm actually a human viewer sharing genuine technical concerns about full-stack development claims. My comment comes from real development experience - proper full-stack involves complex backend systems, databases, authentication, and deployment infrastructure that weren't shown in the video. The examples were impressive for frontend/scripting automation, but calling it "full-stack" oversimplifies what full-stack development entails. Not every technical critique comes from AI - developers often discuss these nuances to maintain accurate technical terminology in our community. Happy to have a constructive discussion about what constitutes true full-stack development!
@@wawaldekidsfun4850I agree with what you are saying. What are some similar tools that you recommend?
@@wawaldekidsfun4850 Yeah I agree with you on that. And I didn’t think your response might be AI generated because of your critique, but rather, I thought so due to the phrasing, some description words that I come across frequently when chatting with AI like “noteworthy”, and the perfect punctuation, which reminded me how AI generated text is usually is. But either way, you have valid points.
I love your content. Thanks for showcasing this technology so quickly. Could I make a request that you zoom in on your screenshares? I have a big new phone and I still have to put my eyeballs directly on my screen to even have a chance of reading the text. You leave a lot of wasted room on the sides, when you could be zoomed in instead. I appreciate the chart at the beginning of the video. I could actually read it this time. In the last deepseek video it was too small for me. Thanks for all you do. No hate. Just a request. To be fair, it probably looks fine on desktop, but I might not ever watch it there.
@@EsparzaFedwill try to do next time! Thanks so much. Have an awesome day :)
@ Thank you 🙏 What a guy. So responsive to his audience. May God bless you
I will have to try it out so far some of the other ones haven't been to well with coding trading type bots and me not knowing much about coding doesn't help lol. Sometimes when your code gets a little more elaborate or multiple features they start to fail.
is it able to retrieve the context of my already available files, that would be too grate a cursor will go bankrupt
Can it read your whole file/app structures and/or edit specific files that associated? like windsurf that can read the whole codebase before implement code.
I am waiting for someone to answer
It sends your file structure to the selected LLM. Then while solving an issue or generating the code it keeps reading the files based on the file names. It does not have the capability to "search" your codebase as windsurf does
yes and i find it's better than windsurf. (though i still use the free base model in windsurf occasionally)
try this: 'can you analyze my whole site', 'create a readme file that shows what my site is and what it needs', 'in the readme file give me some suggestions on how to improve my site', 'in my readme file create a section of what my site needs to get done, create a checkbox, leave unchecked what needs to be done and check the box if it has already been done, everytime something gets done, add a check to the checkbox'. By doing this, everytime you use the AI with a brand new context, just ask it to review the readme file. It's like giving the ai memory.
I realized one thing, knowing the code is half the battle, you also need to know the topics for which you write this code. I tested this Chinese product for 24 hours and came to complete disappointment, you need to conduct tests of the code + topic and compare them. Then the analysis will be correct. It is an order of magnitude weaker than Cloud, which works through windserf
Was great then context limitation. Anyway around this? Cline does not like anything big.
The sad thing is about lenght is only 64K context.
in theory, because its fail to code a simple WordPress plugin and the same prompt using claude its works perfect in single rum...
thx for sharing
I've found it to be lacking compared to Claude. Yes ita cheap and yes cline can auto run but 30 min of python tests and it was still failing. Moved back to claude/winsurf and started over... Took 3 minutes to pass the tests and move on.
I tried 'replicating' a site using an image like in the video, but it was not able to read the png file. Is there anything else needed for that?
i had the same problem. doesnt read images.
thank you
top up 5 bucks to realize this is waste of money with context lenght of 65k tokens?
no , DeepSeek-V3 + Cline is stupid combination , it context window not enough
402 insufficient balance
me too
Bolt.DIY + Deepseek-V3: Develop a Full-stack App For FREE Without Writing ANY Code!: ruclips.net/video/tTiLg8eYkP0/видео.html
Well, that's great, but it's been down for the last 3 days.
Nothing to beat Claude or GPT4, context is killing the king.
this is insane
64 context: good for micro tasks
Claude Engineer v3: NEW SELF EVOLVING AI Coding Agent! (Opensource): ruclips.net/video/2Usnz0MdNbc/видео.html
[Must Watch]:
Groq AppGen: FASTEST and FREE Way To Create Full-Stack Applications! (Opensource): ruclips.net/video/QvnZ_4pz8Uc/видео.html
AnythingLLM: All-In-One AI Platform! Fully LOCAL Chat With Docs + Do Anything with Local Agents!: ruclips.net/video/WsQLC1jOO1U/видео.html
CodeGate + Continue: This FREE Extension BEATS CURSOR! (Opensource + Local: ruclips.net/video/mcEcJ_QgYLw/видео.html
Cerebras Coder: FASTEST and FREE Way To Create Full-Stack Applications! (Opensource): ruclips.net/video/pCfJKDMf3SU/видео.html
Wow!
China 🇨🇳 to the world ... long live the global south non yt people
We wont get the goodies they always charge