Timestamps:
00:00 - Intro 01:17 - Thanks for 50K Subs 01:40 - First Look 02:42 - Technical Look 04:06 - Local Testing Setup 04:58 - E4B Browser OS Test 08:16 -
E4B Result Improvement Test 09:41 - E2B Browser OS Test 12:29 - E4B Subway Scene Test 14:31 - E2B Subway Scene Test 15:29 - E4B 3D Driving Game Test
16:00 - A Speechless Result 17:29 - E2B 3D Driving Game Test 18:52 - E4B Multimodal Wireframe Site Test 20:39 - E2B Multimodal Wireframe Site Test
21:47 - Native Audio Capability Demo 24:04 - E2B Multimodal Test 24:38 - E4B Multimodal Test 25:26 - Results Overview 27:29 - Android Phone Model
Benchmarks 28:10 - Mobile Phone Use Test Results 29:33 - Closing Thoughts
AI Integration & Consulting: https://bijanbowen.com/ Join the Discord:
In this video, we take a hands-on look at the Gemma 4 small models, testing both the E2B and E4B variants across a wide range of real-world tasks. The
goal is to evaluate how these smaller models perform in local environments and whether they can deliver strong results despite their size.
We begin with a technical overview and local setup, then move into practical tests including browser-based workflows, simulation tasks, multimodal
website generation, audio capabilities, and mobile performance benchmarks. Each model is tested side-by-side to highlight differences in performance,
capability, and consistency.