GosuCoder - Videos
Back to Channel
Claude Max just got more valuable? Claude Code + RooCode
Roo Code and Cline integrated your Claude Max subscription, but how well does it actually work. Links: π§βπ»My Recommended AI Engineer course is Scrimba: https://scrimba.com/the-ai-engineer-path-c02...
Is this a new Claude Code Competitor or an AI Terminal? Warp Dev
Warp.dev is an interesting application, its a terminal but not a terminal, but it also has agentic coding capabilities that are pretty good. Its still early and some parts are a bit broken, but i'm...
Why even say this? Codebase Indexingβ¦
How codebase indexing became controversial is confusing to me, in this video I discuss the different ways AI coding agents navigate codebases, whether they use an index or not. I get into some of t...
Augment Code: The Upgrade You've Been Waiting For
Task Lists in AI coding assistants are so good. Augment Code just updated theirs with a built in task list and its freaking amazing. Links: π§βπ»My Recommended AI Engineer course is Scrimba: https:/...
o3 as my primary AI coding model?
OpenAI's o3 shocked me when it dropped its pricing by 80%, this all aligned with the release of o3-pro. With this latest price drop I got to wondering if o3 could be a daily coding model. Links: π§...
You Aren't Using Claude Code to Its Full Potential...
Claude Code has so much potential, in this video I discuss 15 tips based on how I use Claude Code. Some are pretty sweet, some you may know already. But overall Claude Code is an amazing coding too...
I spent 2 days comparing Jules and Codex...
I spent 2 days comparing like for like tasks between Jules and Codex. In this video I discuss my findings of two of the most popular coding remote agents. Links: π§βπ»My Recommended AI Engineer cour...
WTF Apple why release this AI white paper?
I feel like this whitepaper called the Illusion of thinking: Understanding the Strengths and Limitations of Reasoning via the Lens of Problem Complexity is really stating much of the obvious about ...
Did Google Fix Gemini 2.5 Pro?
Gemini 2.5 Pro is thankfully fixed, I hope they donβt regress it again. Massively improved tool calling, and much better quality code at 0.7 temp. Links: π§βπ»My Recommended AI Engineer course is Sc...
Claude Code is so freaking good
I spent $200 for Claude Max and have zero regrets. Claude Code really is just a ton of fun to use. Timestamps 00:00 Claude Code May results 01:06 OS support and Windows/WSL setup 02:37 Pricing: Pr...
I think now is the BEST time to learn to code! Don't Fear the AI
There is so much media around why we shouldn't learn to code in 2025, that I decided to share where I've been landing as I've thought through the possible outcomes. I think now is probably one of ...
The BEST AI Coding Assistants | June 2025
I built a new set of evals that specifically measure how a model performs with an AI Agent. The results of these change month over month so I want to test out starting a series to see how things ch...
DeepSeek just improved R1, does it work better in AI Coding Tools?
DeepSeek R1 05-28 is a solid upgrade over the previous version, but still not a good model as your main coding model due to a few limitations. My Links π ππ» Subscribe: https://www.youtube.com/@G...
I've been waiting for this feature | full codebase indexing in RooCode
While these new features are experimental, its incredibly exciting to see the progress being made here. In my tests I found RooCodes Semantic search paired with Claude 4 did a comparable job to Aug...
This is currently the best local model for RooCode | Devstral
I've been on the search for a local model that works well with RooCode, and Devstral finally made it a reality. Fresh off a new install of my RTX 5090 I put Devstral through my evals which is the o...
Claude 4 comes with big updates for AI Assisted Coding & New Evals ran!
Claude 4 announced today, and I got right to work testing it. Surprisingly the Thinking version enabled in RooCode took the top spot. I do think the addition of MCP connections directly in the api ...
Claude 4 Today? Live Stream
Anthropic's developer conference, hoping for a Claude 4 release!!!
Gemini 2.5 Flash got a massive upgrade!
Gemini 2.5 Flash announced as an upgrade at Google I/O, so I had to set out to test. My Links π ππ» Subscribe: https://www.youtube.com/@GosuCoder ππ» Twitter/X: https://x.com/GosuCoder ππ» LinkedI...
AI Assisted Programming keeps getting better | Codex review after 20 hours of usage
OpenAI's codex is absolutely incredible, and at $200 a month its definitely worth it if you have the work to send through it. My Links π ππ» Subscribe: https://www.youtube.com/@GosuCoder ππ» Twit...
Windsurf's new AI Models are ... interesting | SWE-1 Fully Tested
Windsurf surprised everyone by launching their own series of models. In this video we run the models through an entire day of coding, as well as Evals to see how they stack up. I'm impressed with h...
Scaling AI without a Massive Budget: DeepSeek V3 is a Marvel
DeepSeek V3 Whitepaper: Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures : https://www.arxiv.org/pdf/2505.09343 This makes me very excited for DeepSee...
Is Zed IDE the "Next Gen Code Editor" It Claims to Be?
The new Zed IDE with agentic coding has some really good features, but I do wonder if it can eat into the behemoth that is VS Code and their extensions. And its OpenSource which is incredible. My ...
Remote Agents will change AI Assisted Programming Forever
Augment Code recently announced Remote Agents, which I believe others will follow with very quickly. In this video I talk about how I think this will change AI Assisted programming forever. My Lin...
Gemini 2.5 Pro Just Became the #1 Coding LLM - But it needs some tweaking
Gemini 2.5 Pro 05-06 Update is a massive boost in coding My Links π ππ» Subscribe: https://www.youtube.com/@GosuCoder ππ» Twitter/X: https://x.com/GosuCoder ππ» LinkedIn: https://www.linkedin.com...
GitHub Copilot keeps getting worse. Please turn this around MS.
GitHub Copilot went from being one of my daily use AI Coding Assistants to unusable, and I think its about to get worse. Timestamps: 0:00 Copilot Problems 0:56 Bugs & Glitches 2:40 Painfully ...
Augment Code Releases Fair Pricing Details | They Listened to Us
Augment Code is without a doubt one of the best AI Coding Assistants on the market. This latest change is so exciting, because it shows unlike other companies in the space, they listen and implemen...
Anthropic's Brilliant Move - Claude Code with Max Subscription
Anthropic's made a brilliant strategic play by letting users get nearly infinite Claude Code usage for $200 a month with a Claude Max sub. This is something we need to see other companies do ASAP. ...
This AI Model has me excited about the future of Local LLM's | Qwen3-30B-A3B
Qwen3-30B-A3B may not be the best model at coding, but its ability to run incredibly fast on modest hardware has me excited about the future of Local MoE models. Especially ones geared towards codi...