New Ollama Update Is INSANE!

@jgoldieseo

15.0K followers · usually 1.8K views a post

Views

6.8K

Likes

220

Comments

3

Shares

33

Selling · SaaS › AI Tool

Ollamaseen in 8 videos →https://www.skool.com/ai-profit-lab-7462/about ↗

Why it worked

The video effectively demonstrates a significant improvement in local AI capabilities, addressing a common pain point for users. By showcasing concrete performance gains and speed increases, it appeals to users seeking powerful AI solutions on their own hardware.

Summary

This video highlights the significant improvements in the latest Ollama update, particularly its enhanced tool-calling capabilities for Gemma 4, which now performs much better on benchmarks and runs faster on local machines using Apple's MLX engine. The update allows local AI models to complete tasks more reliably, closing the gap with cloud-based solutions.

Structure

1Introduction to the Ollama update
2Demonstration of local AI limitations
3Highlighting Gemma 4's improved performance
4Explaining the speed increase with MLX engine
5Concluding on the reliability of local AI

Product placement

Original caption

Local AI models used to quit halfway through tasks. This update changes that 🧠 Gemma 4's tool-use score jumped from 6.6% to 86.4% on benchmark testing. Ollama's latest update brings that tool calling to models running fully on your own hardware. Nothing gets sent to the cloud. Local AI just closed a big part of the gap 👇 #ollama #localai #gemma #aitools #opensourceai

More from @jgoldieseo

Their other posts in the index, biggest breakout first.

Ollama (software): enables running AI models locally on a user's machine. It ACTS by processing AI tasks and improving performance. Removing it would change what happens in the video. Gemma 4 (AI model): an AI model that is being tested for its tool-calling capabilities. It ACTS by performing tasks and being benchmarked. Removing it would change what happens in the video. Gemma 3 (AI model): an AI model that was previously benchmarked. It ACTS by performing tasks and being benchmarked. Removing it would change what happens in the video. MLX (engine): Apple's engine that speeds up AI model performance on Apple Silicon. It ACTS by improving the speed of AI models. Removing it would change what happens in the video. AI Profit Boardroom (course/community): a platform for learning about AI and profit strategies. It APPEARS as the creator's bio link destination. Removing it would not change what happens in the video.

On-screen text

New Ollama Update Is INSANE! NEW OLLAMA UPDATE THIS NEW OLLAMA UPDATE MAKES AI ACTUALLY FINISH THE LOCAL AI MODELS WERE SMART DO ANYTHING TASK OR FORGET OR CLAIM THEY FINISHED SOMETHING THEY OLLAMA'S NEW UPDATE FIXES THAT WITH FOR GEMMA 4 A REAL BENCHMARK TESTING HOW WELL AI USES TOOLS SCORED 6.6% SCORED 86.4% FASTER TOO APPLE'S MLX ENGINE PUSHED SPEEDS FROM 50 TO 95 TOKENS YOUR OWN MACHINE LOCAL AI JUST CAUGHT UP

Transcript

New Ollama update is insane! This new Ollama update makes AI actually finish the local AI models were smart do anything task or forget or claim they finished something they Ollama's new update fixes that with for Gemma 4 a real benchmark testing how well AI uses tools Gemma 3 scored 6.6% Gemma 4 scored 86.4% faster too Apple's MLX engine pushed speeds from 50 to 95 tokens your own machine local AI just caught up