@rpn
Instagram · Greenscreen Talking Head
@rpn You want to use DeepSeek R1 but you don’t want your data processed in China? Run it locally 🫡 Here’s an easy way to install and run DeepSeek R1 or Llama locally. This will let you: - save money - protect your data - use them offline - avoid censorship Open-source is the future of AI. ✌️ #deepseek #llama #ai
Performance
Video #14452 · Greenscreen Talking Head
[01]
The creator provides a four-step tutorial on how to install and run open-source AI models like DeepSeek and Llama locally. He walks through downloading Ollama, using terminal commands to install a model, downloading the Chatbox AI client, and configuring it to run the local model, while showing screen recordings of each step.
[02]
[03]
622.5KSplit screen AI news explainer@eluna.ai · Explainer
503.3K17.3×Creator gives 5-step AI setup guide@itsmariahbrunner · Tutorial / How To
144.2K3.8×Tutorial for AI app self-improvement@angus.sewell · Tutorial / How To
125.9K4.3×Use AI scraping to learn@angus.sewell · Tutorial / How To
206.2K11×Tutorial for creating a personal AI@angus.sewell · Tutorial / How To
273.3K30.5×Explaining new open-source AI tool@brodyautomates · Opportunity Explainer
224.9KAI tool automates Instagram tasks@taki.gpt · Product Demo
541.5K1.1×AI builds a castle in Blender@rileybrown.ai · Tutorial / How To
596.8KClaude AI Instagram audit tutorial@laurahaleydt · Tutorial / How To
193.8K30.2×AI call agent product demo@jasoncooperson · Product Demo
649.3K6.3×Creator explains AI mom's system@rony · Case Study Breakdown
4M647.5×AI app builder product demo@lovable.app · Product Demo[04]
[05]
What stops the scroll in the first 3 seconds
Title hook
show you an
Verbal hook · 65%
Gonna show you an easy way to install DeepSeq or llama locally on your machine.
Visual hook · 35%
A split-screen layout with a speaker on the bottom and a screen recording of an app store on the top, set against a dynamic background of a whale shark swimming underwater.
[06]
Steal this for your own video
I’m gonna show you the easy way to on your .
[07]
The promise the opener makes, and how the video pays it back
Hook
State a clear value proposition by offering an 'easy way' to achieve a desirable, complex goal, immediately followed by listing the key benefits (e.g., saving money, gaining control, offline access).
Payoff
Provide a 'Proof of Success' by demonstrating the fully functional end result in real-time. This is immediately followed by a 'Pattern-Interrupt Outro' where the creator abruptly shifts to a personal, unrelated opinion to humanize the content and spark off-topic engagement.
[08]
The structural pattern this video runs on
Framework
The Step-by-Step Empowerment Guide
A tutorial that breaks down a complex technical process into simple, numbered steps, culminating in a functional demonstration and ending with a personal, off-topic signature to build community.
Step 1: The Foundational Tool
Instruct the viewer to acquire the first, most fundamental component required for the process. This serves as a low-friction starting point.
Step 2: The Core Configuration
Guide the viewer through the most technical step using a command-line or code-based interface, providing the exact inputs visually to minimize error and build a sense of technical accomplishment.
The Technical Clarification
Briefly pause the tutorial to explain a key variable or concept within the technical step, adding a layer of educational value and demonstrating expertise.
Step 3: The Interface Layer
Instruct the viewer to acquire a second, user-friendly tool that will act as the interface for the core component, bridging the gap from technical to practical.
Step 4: The Final Integration
Show the final configuration step, connecting the foundational tool with the interface layer within the application's settings to complete the setup.
You want to use DeepSeek R1 but you don’t want your data processed in China? Run it locally 🫡 Here’s an easy way to install and run DeepSeek R1 or Llama locally. This will let you: - save money - protect your data - use them offline - avoid censorship Open-source is the future of AI. ✌️ #deepseek #llama #ai
title · middle
DeepSeek R1
informational · middle
ollama run deepseek-r1:1.5b
title · middle
Llama 3.3
informational · middle
ollama run llama3.3
title · middle
Llama 3.2
informational · middle
ollama run llama3.2:1b
Gonna show you an easy way to install DeepSeq or llama locally on your machine. That way you can save money, use them offline, and have complete control over your data. So step one, head over to ollama.com and download it. Step two, once you've installed it, open the terminal and type in one of the commands in yellow for deep seek and these are the ones for llama. Now the number at the end refers to the weight of the model. The heavier the model, the the more accurate and better the performance, but the more powerful your computer needs to be. So for this example, I'm installing the lightest 1,500,000,000 parameter version of DeepSeek, which is about a gig in size. Step three, head over to chatboxai.app, download it, and install it. Step four, once you've opened Chatbox, select ollama from the provider drop down and choose the DeepSeek model that we just installed. And now you're running DeepSeek locally, and you can see how it thinks and reasons in real tier, which is eerily human. Hope that's helpful. And, by the way, I got a feeling Mahomes is gonna win 10 rings. Playoff bad is unstoppable. Dude's a menace. I'll see you tomorrow.