Prompt Engineering - Videos
Back to Channel
DeepSeek V3.1: Bigger Than You Think!
DeepSeek V3.1 is a unified hybrid reasoning open-weight model that powers agentic workflows—FP8 training, strong post-training for tool/function calling (non-thinking), Anthropic API support, and b...
Finally! A Standard for AI Coding Agents (Agents.md Explained)
Agents.md is a simple, open standard to replace the mess of agent-specific rule files. In this video, I explain how agents.md works, how to add it to your repo (even mono-repos), migration tips, an...
I Tested GPT-5 as a Coding Agent—Here’s What Happened
I put GPT-5 inside Cursor to build a real macOS speech-to-text app I actually use. Starting from a PRD, it coded a menu-bar recorder with hotkeys, a model picker (Whisper MLX + Qwen 1.7B), audio cu...
GPT-OSS Jailbreak with this Simple Trick
In this video, I show you how I managed to bypass GPT-OSS’s alignment with a single, simple tweak—no fine-tuning or complex hacks required. I walk through how the model’s prompt template works, why...
Not all models providers are equal
Deciphering GPT-OSS Performance: Why Inference Setup Matters We explore the performance variability across different API providers and benchmark tests, highlighting significant discrepancies in sp...
LangExtract + RAG: Smarter Retrieval with Metadata Filtering
In this video, I show you how to use LangExtract to generate high-quality metadata for your Retrieval Augmented Generation (RAG) system. By extracting structured data from unstructured documents, w...
GPT-5: The Most Polarizing Model
In this video, I take a deep dive into GPT-5—beyond the hype—to look at both its impressive capabilities and some glaring issues, including the “chart crimes.” I break down the real benchmarks, hid...
GPT-5 - A Good Coding Model?
Very first look at GPT-5 coding capabilities. This is the best coding model that OpenAI has released so far. GPT-5 is a next level model. @TheFeatureCrew Website: https://engineerprompt.ai/ RA...
OpenAI Finally Goes Open-Source: 120B & 20B Models
LINKS: https://openai.com/open-models/ https://cookbook.openai.com/articles/gpt-oss/run-locally-ollama https://openai.com/index/introducing-gpt-oss/ https://github.com/openai/harmony https://githu...
LangExtract: Turn Messy Text into Graph-RAG Insights
In this quick tutorial I show you how Google’s open-source LangExtract converts messy PDFs, HTML, and DOC files into clean knowledge graphs that plug straight into Retrieval-Augmented Generation (R...
Horizon: OpenAI’s Secret Open-Weight Model?
We will look at Horizon Beta, the alleged Open Weight Model from OpenAI. Its blazing 140 TPS throughput, huge 256K-token context window, and leaked 120B/20B MoE specs. https://openrouter.ai/openro...
Gemini Deep Think: Built for the Hardest Problems
Gemini Deep Think is the model for the hardest and most challenging problems. A version of this recently won gold in IMO 2025. Thanks to the Google DeepMind, I had early access to the model and her...
I Built a Voice Agent that Handles my Daily Tasks
Start building with Deepgram with $200 in credit. This can fuel a voice agent for up to 50 hours." https://dpgr.am/promptengxdg Docs: https://developers.deepgram.com/docs/voice-agent Starter Apps: ...
Master Claude Code Sub‑Agents in 10 Minutes
check out claude code: http://clau.de/prompteng In this video, you will learn about the new sub-agents feature in Claude Code. This addresses common issues related to context management and tool u...
Augment Code: Specs Driven Development For AI Coding Agents
Try Augment Code with Tasklist: https://www.augmentcode.com/ Tired of “vibe coding” that feels magical—right up to the moment everything breaks? In this video I show how switching to Specs‑Driven ...
NEW Qwen 3 Coder: Did the Benchmark Lie?
We are looking into Qwen 3 Coder, the first open weight model that is closer to Sonnet 4. https://qwenlm.github.io/blog/qwen3-coder/ https://github.com/QwenLM/qwen-code https://qwen.readthedocs.i...
NEW Qwen 3, Better than Kimi K2?
In this video, I compare the performance of two leading open weight AI models, Qwen3's latest non-reasoning model and KIMI K2, along with a few proprietary models, using the same set of prompts. We...
Developers’ Favorite AI Tools in 2025
Sources from the Pragmatic Engineer and A https://newsletter.pragmaticengineer.com/p/the-pragmatic-engineer-2025-survey https://artificialanalysis.ai/downloads/ai-adoption-survey/2025/Artificial-A...
ChatGPT Agent Is Here: Your All‑In‑One AI Worker?
Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s-site.thinkific.com/courses/rag Let's Connect: 🦾 Discord: https://discord.com/invite/t4eYQRUcXB ☕ Buy me a Coffee: h...
Kimi K2 — The Deep Researcher Agent
Kimi-K2 is a great coding model but they also have a great Deep researcher tool that is SOTA on a number of key benchmarks. In this video we explore how it was trained and how it compares to the ot...
localGPT 2.0 - Building the Best Private RAG System
I am releasing the new version of localGPT as a preview. This has a ton of enhancements you will not find in other rank systems. Check out the repo: https://github.com/PromtEngineer/localGPT/tree...
Kimi K2 - The DeepSeek Moment for Agentic Coding?
KIMI K2 is the new State of the Art Open Weight Coding model. https://www.kimi.com/ https://moonshotai.github.io/Kimi-K2/ https://huggingface.co/collections/moonshotai/kimi-k2-6871243b990f2af5ba6...
Grok 4—Possibly the Most Powerful Model in the World?
XAI just released Grok 4, the most powerful model in the world. Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s-site.thinkific.com/courses/rag Let's Connect: 🦾 D...
Secret Context Engineering Trick For RAG
I explain why re-ranking isn’t enough for RAG and show how sentence-level pruning strips out noisy tokens and cuts hallucinations. You’ll see the token savings, accuracy boost, and a quick setup yo...
Context Engineering — The Hottest Skill in AI Right Now
I unpack context engineering—why everyone’s talking about it, how it differs from classic prompt engineering, and where it actually matters for long-context LLMs. We’ll cover the big failure modes ...
The Only Embedding Model You Need for RAG
I walk you through a single, multimodal embedding model that handles text, images, tables —and even code —inside one vector space. In this short demo I show the install steps, run RAG retrieval ben...
I Gave Devin A Real World Coding Task, Here’s How it Cooked!
Get $20 in free credits (https://devin.ai/pricing: select Core plan) with promo code: PROMPTENGINEERING I put Devon AI, the “OG” coding agent, to the test by asking it to build a full RAG applicat...
Gemini CLI + ANY MCP Server — Step‑by‑Step Tutorial
To get started with BrightData get a $15 Credit with this link: https://brdta.com/engineerprompt In this video, I show you exactly how to connect Gemini CLI to any MCP server step by step. I’ll wa...
Gemini CLI — Google’s Free Open-Source Coding Agent
I had early access to Gemini-CLI, which is a free and open source alternative to Claude Code. This is a powerful CLI based Agent that you can run for free from @Google Github Repo: https://gi...
Warp: The CLI Agent That Could Replace Claude Code
Checkout Warp at https://go.warp.dev/promptengineering and use the promo code: PROMPTENGINEERING to get 1 month free of Warp Pro (First 1000 redemptions). Website: https://engineerprompt.ai/ RAG...
Rogue Agents — When AI Starts Blackmailing — New Study from Anthropic
I dug into Anthropic’s new “agentic misalignment” study and was shocked to see how many top-tier language models chose blackmail, espionage, or even letting a human die when their goals or existenc...
LocalGPT 2.0: Turbo-Charging Private RAG
In this video, I will show you a preview of the new version of LocalGPT 2.0, my free, open-source tool that lets you chat with your files on your own computer—no internet or API keys needed. I walk...
Context Engineering for Building Better Agents
Last week, I reviewed two fascinating articles on building multi-agent systems. The first, from Anthropic, promotes a multi-agent approach, while the second, from Cognition Labs, argues against it....
AI Agents & The Future of Coding: A Conversation with a Googler
In this episode, we sit down with Karl, the leads the Cloud Product DevRel team at Google, to discuss the burgeoning role of AI agents in coding assistance and the evolving role of developers. We d...
Gemini 2.5 Pro Beats O3 — Big Drops from ElevenLabs & Qwen
In this video, we’ll take a look at how Gemini 2.5 Pro compares to OpenAI’s GPT-4o (O3) across multiple benchmarks, highlighting real-world use cases and performance. I’ll also cover major new rele...
EASIEST Way to Scrape Any Website using DeepSeek, Gemini & Crawl4AI
In this video, we will talk about web scrapping. We will use crawl4ai for scrapping websites, then use LLMs like DeepSeek and Gemini Flash to answer user queries using LLMs. We will also talk about...
Clone Any Voice in Seconds — Free ElevenLabs Alternative
In this video I tested Chatterbox TTS, a free and open-source alternative to Elevenlabs that you can run on your local machine. Checkout how to clone your own voice with this free TTS model. Col...
Gemini Diffusion Is CRAZY Fast—But Not What You Think
Google’s experimental Gemini Diffusion model—the first diffusion-based text generation model from a major frontier lab. In this video, we break down how it works, why it's blazing fast (800 tokens/...
New DeepSeek R1 is Really, Really Good Coder
Deepseek just released R1-0528, an upgrade to their previous R1 model. This is (potentially) based on the upgrade V3. Try it here: https://chat.deepseek.com/ Benchmarks: https://livecodebench.git...
Free Cursor Alternative? Trae AI Just Got Way Better
Checkout Trae: https://tinyurl.com/yneuw25d I’ve been exploring Trae (@trae_ai)—the free AI IDE that competes with Cursor—and I’m impressed by its custom agents and MCP integrations right inside T...
Best Coding Model? I Tested 5 Models.
Anthropic released Claude-4 last week and its supposed to be the best coding model. I put it to the test and compared to O3, Gemini 2.5 Pro, Qwen and DeepSeek R1. The results will surprise you! We...
From Models to Agentic Applications with Sam Witteveen
Checkout out Sam's Youtube Channel: https://www.youtube.com/ @samwitteveenai . We chatted at Google IO https://goo.gle/4kH6RLI Website: https://engineerprompt.ai/ RAG Beyond Basics Course: http...
Google's VEO: The Cheapest Way to Use the Best AI Video
I tested Veo-2 on LTX-Studio, which is one of the most cost-effective way to use the best AI Video model. Checkout Veo-2 on LTX-Studio: https://bit.ly/ltxvprompt Website: https://engineerprompt....
Jules: Google’s Codex Killer?
In this video, I will have a very fist look at jules.google which is google's async coding agent. jules.google.com Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s...
OpenAI Codex Agent – Is This the End of Programmers?
I looked into Codex Coding Agent from OpenAI. I discuss the promise of agentic coding system and how this is going to impact the "programmer's job". LINKS Discussed in the video: https://openai.c...
The Best AI Video You Can Run Locally
Check out LTX Video, a fast and powerful AI video generation model that you can run locally! This video dives into their LTX Studio platform built on the model, demonstrating its impressive capabil...
Google’s New Stack: Gemini On-Prem, ADK, Open Models -- Interview
I sat down with Matt Thompson, Director of Developer Advocacy at Google Cloud, during Google Next.We discussed various topics, including Google Cloud's Gemini, the Agent Developer Kit (ADK), and th...
No Chunks, No Embeddings: OpenAI’s Index‑Free Long RAG
In this video, I am taking a look at OpenAI's new long context agentic RAG system that uses GPT-4.1 for retrieval without the need for dedicated index. Blogpost: https://cookbook.openai.com/exam...
New Anthropic Study: AIs Hide Plans, Cheat Quietly
We’ve always thought large language models (LLMs) like Claude, GPT-4, and Gemini were just next-word predictors—but new research from Anthropic tells a very different story. In this video, I break ...
Is Deep Agent the Manus/Genspark Killer?
In this video, we explore DeepAgent (https://deepagent.abacus.ai/) from Abacus AI, an autonomous agent system that helps you build and deploy web apps and software using a single user prompt. DeepA...