Video
Fahd Mirza
2026-09-23
Opus 5.5 vs GPT-6 Sol: Can AI Actually Simulate a Latte?
Compares Claude Opus 5.5 and GPT-6 Sol performance simulating a latte.
Videos
Videos about AI and LLM
Find videos about LLM models, hardware, runtimes, benchmarks and practical AI workflows.
1083 videos found
Showing 251-300
Video
Bijan Bowen
2026-09-22
Claude Opus 5.5 Is INSANE – Hands-On With the BEST Model Yet!
Tests Claude Opus 5.5 via browser OS, C++, Blender, Godot, Subway FPS, and an original benchmark.
Video
WorldofAI
2026-09-22
Claude Opus 5.5 IS THE Greatest AI Model EVER! Cheaper, Fast, & Powerful! (FULLY TESTED)
Benchmarks Claude Opus 5.5 via woaibench.ai across agentic coding, complex frontend development, visual design, games, and real-world tasks, evaluating performance, speed, and efficiency per task.
Video
Alex Ziskind
2026-09-22
M5 Ultra… Apple Wasn’t Messing Around
Evaluates the performance requirements needed for the M5 Ultra chip to achieve its full potential.
Video
Luke's Dev Lab
2026-09-22
DavidAU Qwen3.8 27B TURBO Fable Cold Fusion tested - 16GB Local LLM setup
Tests performance, memory, reasoning, OpenAI Human Eval, Kanban, Sand Physics, Dungeon Crawler, Blender, and Godot using the DavidAU Qwen3.8 27B TURBO Fable Cold Fusion GGUF model locally.
Video
No place like localhost
2026-09-22
Occamy 1.0 - first look and test
Evaluates Occamy 1.0, a model post-trained on Qwen3.6 35B A3B, against its base model using tts-serve to verify performance claims through multiple attempts and code review.
Video
Execute Automation
2026-09-22
Mac mini M6 First Look — Qwen3.8-27B Hits 53+ TPS Locally
Tests Qwen3.8-27B local inference on the Apple Mac mini M6, achieving 53+ TPS while evaluating M6 chip features including CPU, GPU, unified memory, and Neural Accelerators.
Video
WorldofAI
2026-09-22
Xiaomi MiMo-V2.6 Pro IS THE BEST Open Source Model EVER! (Fully Tested)
Tests Xiaomi MiMo-V2.6 Pro across coding, frontend design, 3D game development, multimodal tasks, and computer use, comparing performance against other frontier AI models.
Video
Fahd Mirza
2026-09-22
Xiaomi Launches MiMo-V2.6-Pro as Top Open AI Model: Let's Verify
Tests MiMo-V2.6-Pro-RL, an open AI model designed to scale reinforcement learning toward self-improvement.
Video
AI Search
2026-09-22
Finally! New best local AI image editor is here
Review and tutorial covering installation, text-to-image generation, image editing, background removal, LoRA loading, and low VRAM setup for Qwen Image 2.1 within ComfyUI.
Video
Bijan Bowen
2026-09-21
Grok 4.7 Is HERE – Is THIS the CHEAPEST Frontier Model?
Hands-on test of Grok 4.7 across browser workflows, C++, robot arms, Chrono City, Blender, Godot, and FPS tasks to evaluate its performance against costs.
Video
Fahd Mirza
2026-09-21
Grok 4.7: I Gave It a Broken Championship, a Broken LED, and Coffee
Evaluates Grok 4.7, described as xAI's most capable model for coding and knowledge work, using tasks involving a broken championship, a broken LED, and coffee.
Video
KGP Talkie
2026-09-21
Qwen Image 2.1 Low VRAM Test: 4-bit Runs Under 16 GB of RAM
Runs Qwen Image 2.1 locally on an RTX 5090 using fp8 and NF4 precisions, testing VRAM usage, render times across resolutions, and cost against GPT Image 2 and Nano Banana Pro.
Video
Codacus
2026-09-21
Can You Run Any LLM in Jev Mode Using llama.cpp?
Demonstrates a /v1/decision endpoint on llama-server using constrained choice, achieving 17.3 ms per decision on a 2 GB model in 2.4 GB VRAM versus 3,415 ms for JSON generation.
Video
Execute Automation
2026-09-21
Qwen-Image-2.1 on M5 Max — Multi-Image Editing, Transparency & Realism
Runs Qwen-Image-2.1 locally on Apple M5 Max via Python to demonstrate multi-reference generation, image editing, transparent RGBA output, identity preservation, typography, and realistic portraits.
Video
Cole Medin
2026-09-21
Jev is the FIRST of a Whole New Class of AI Models (Here's How to Actually Use It)
Demonstrates Jev making probabilistic decisions in ~200ms for under a cent per query, applied to gaming, pull request triage, and agent model routing within an AI coding setup.
Video
xCreate
2026-09-21
Let's Run the LEAKED Step-5 Locally (New #1 Open Source AI) 🤯
Runs Step-5 Preview locally using Inferencer v2.4.1 on Mac Studio M3 Ultra 512 to evaluate its performance against claims of topping open weight charts.
Video
Fahd Mirza
2026-09-21
Qwen-Image 2.1 Hands-On Locally: Prepare a Royal Paan with AI
Local installation and testing of Qwen-Image-2.1, a unified text-to-image generation and image editing model, used to prepare a royal paan.
Video
AI with Eric
2026-09-20
I Got Early Access to Qwen-Image-2.1 - Amazing LOCAL AI Image Model
Hands-on comparison of Qwen-Image-2.1 against Krea-2 and ChatGPT gpt-image-1.5 (2?), evaluating image quality, text-to-image generation, and image editing capabilities.
Video
Fahd Mirza
2026-09-20
Meet Qwen3.8-LiveTranslate: Qwen's Real-Time Live Interpretation Model
Tests Qwen3.8-LiveTranslate, evaluating its ability to name the speaker while carrying the meaning during real-time live interpretation.
Video
Fahd Mirza
2026-09-20
Alibaba Open-Sourced Model That Detects Cancer From CT Scans - I Ran It Locally
Local installation and testing of RADAR, an Alibaba open-source medical AI model capable of detecting cancer and nearly 150 other conditions from CT scans.
Video
AI Search
2026-09-20
OpenAI hacked, Jev, Google’s RSI, Qwen 3.8 Omni, Bonsai 2, new Gemini Live: AI NEWS
Covers Jev, Laya, Bonsai 2 27B, Dream RSI, Gemini 3.8 Live, Qwen 3.8 Omni Flash, Qwen 3.8 Live Translate, Needle 3, ZAI infra, Xiaomi realtime RL, and an OpenAI hack report.
Video
AI with Eric
2026-09-20
Bonsai 2 Claims 98.2% of Qwen. Coding Benchmarks Tell Another Story
Evaluates Bonsai-2-27B against Qwen3.8-27B using Terminal-Bench 2.1 and SWE-bench Verified across automation, orchestration, sysadmin, cybersecurity, and complex coding tasks.
Video
Fahd Mirza
2026-09-19
CUA S1 Forms: Jev-Like for GUI Form Filling Model Locally
Locally installs and tests cua-s1-forms, a small, jev-like one-pass option scorer for GUI form filling.
Video
Execute Automation
2026-09-19
Bonsai 2 27B: Qwen3.8-27B Compressed to Just 6GB — But Is It Still Good?
Bonsai 2 27B compresses Qwen3.8-27B from ~54 GB FP16 to ~6 GB using ternary weights, retaining ~98.2% of aggregate benchmark performance per PrismML.
Video
Fahd Mirza
2026-09-19
Realtime Venus Running Locally: A Full-Duplex Interaction System
Locally installed Realtime-Venus demonstrates proactive audio-visual interaction, asynchronous delegation, and interruption-aware full-duplex dialogue capabilities.
Video
Jose Romero
2026-09-19
Qwen 3.8 Flash-Next: Is Halogen Worth the Switch?
Qwen 3.8 Flash-Next ran on Halogen Flash Server via 72 coding jobs across OpenCode, dsh, and Pi harnesses on a 128 GB Strix Halo system.
Video
Unbiased Bob
2026-09-19
i ran bonsai 2 for 24 hours so you don't have to
Compares Bonsai 2 against Qwen 3.5 9B, Qwen 3.6 35b-a3b and Bonsai 1 using Bob-Bench scores on 8GB VRAM configurations.
Video
Fahd Mirza
2026-09-18
Does Swift Qwen3.8 27B Really Think Less? Comparison with Original
Installs and compares Swift-Qwen3.8-27B with Original Qwen to evaluate whether the claim of fewer thought tokens holds true.
Video
xCreate
2026-09-18
How to Run Claude Code & OpenCode FREE Locally? In-Depth Review 🤔
Evaluates Claude Code, OpenCode and native Inferencer using Qwen4 Exp on Mac Studio M3 Ultra via Tic-Tac-Toe generation, modifications, image inferencing and token usage comparisons.
Video
Alex Ziskind
2026-09-18
Llama Cpp Flags That Instantly Speed It Up
Demonstrates configuring llama.cpp flags to achieve two to three times faster local large language model inference performance.
Video
No place like localhost
2026-09-18
Spec-driven vibe coding in action
Demonstrates combining vibe coding with rigorous specifications through iterative refinement passes before implementing the Dust to Dominion game using an LLM to generate 100% of the code.
Video
Matt Wolfe
2026-09-18
AI News: All AI Labs Want To Slow Down (Except One)
Covers AI pacing debate, Claude Cowork merge, Gemini Notebook study tools, Siri AI launch, Qwen3.8 Omni Flash, Meta One subscription, and UBTECH robot factory updates.
Video
Execute Automation
2026-09-18
Qwen3.8-Flash-Next Hits 109 TPS on M5 Max — MTPLX v2.11.3
Tests Qwen3.8-Flash-Next on M5 Max using MTPLX v2.11.3, which fixes 8 defects in decoding, caching, tokenization, sampling, state restoration, and tool handling.
Video
Luke's Dev Lab
2026-09-18
Moonshot AI - Kimi K3 tested
Tests Kimi K3 across Performance, Memory, Agency, OpenAI Human Eval, Kanban, Sand Physics, Dungeon Crawler, Blender and Godot tasks.
Video
Fahd Mirza
2026-09-18
Bonsai 2 27B Locally: Near-Lossless Compression in a 9x Smaller
Locally installs and tests Ternary-Bonsai-2-27B-gguf.
Video
Jose Romero
2026-09-18
Running Qwen 3.8 Flash-Next at xhigh on 3 harnesses
Tests Qwen 3.8 Flash-Next generation speed across OpenCode, Pi and DeepSeek Harness using 18 benchmark jobs at xhigh thinking level on a GMKtec EVO-X2 with Strix Halo hardware.
Video
Bijan Bowen
2026-09-18
Bonsai 2 27B First Test – Is THIS the BEST Single-GPU AI Model?
Tests Bonsai 2 27B across browser workflows, website generation, C++ game development, Blender scenes, FPS games, frontend design, image-to-SVG conversion, and simple 3D games.
Video
AI Search
2026-09-18
Deepseek just did the impossible
Explains Deepseek V4.1 Flash architecture covering prefill versus decode, causal encoder decoder, global versus local cache, sliding window attention, CSA2, hierarchical sparse indexer, single pass mH
Video
Fahd Mirza
2026-09-17
Qwen3.8 Omni Flash: Fed It Images, Video, and Audio All at Once
Tests Qwen3.8-Omni-Flash, a native omnimodal model, using images, video, and audio inputs simultaneously.
Video
Zero to MVP
2026-09-17
What Local AI Models Can a Regular Computer Run?
Token generation speeds were measured for Gemma 4 E2B, Gemma 4 12B, gpt-oss 20B, Gemma 4 26B, Qwen 3.6 35B, and Qwen 3.8 27B using Ollama on a Geekom A9 Max mini PC with 32 GB RAM.
Video
Execute Automation
2026-09-17
Meet Spark-X2.5-4B — A Tiny 4B Model with 1M Context & Agentic Coding
Tests Spark-X2.5-4B agentic coding via Playwright MCP tool calling and .NET 10 migration, highlighting its hybrid attention architecture combining 3 sliding-window and 1 full-attention layers.
Video
Fahd Mirza
2026-09-17
Tencent Shrank a 1.5TB AI Model to 214GB — Here's the Trick
Explains how Tencent used Sherry quantization to shrink a 770B model from 1.5TB to 214GB.
Video
Cole Medin
2026-09-17
You Built Your AI Second Brain. Now What? (Here's How to Evolve It)
Demonstrates a team brain architecture using Oracle AI Database, consolidating data from chats, repos, and docs into one table with labeling, row-level security, and MCP access for coding agents.
Video
Fahd Mirza
2026-09-16
Union Alpha Stealth Model Tested: Coding, Vision, Reasoning
Tests Union Alpha, a multimodal model designed for research, coding, and agentic workflows.
Video
Lukasz Gawenda
2026-09-16
Official SGLang Beat My Qwen3.8 Flash Next Build by 56%, Then I Ran Real Tests!
Tests Qwen3.8-Flash-Next on RTX PRO 6000 with SGLang, measuring speed, BFCL, tau^2-bench, long context, strategic reasoning, SVG drawing, video editing and motion design capabilities.
Video
Alex Ziskind
2026-09-16
8 RTX Pro 6000’s Wasn’t What I Expected
Tests eight RTX Pro 6000 GPUs with 768GB VRAM installed in a Comino Grando chassis to evaluate performance limits.
Video
No place like localhost
2026-09-16
Running AI group chat in very low VRAM with LuxTTS
Demonstrates running TalkWithMe v7.1 with LuxTTS, tts.serve v1.1, optimized STT and LLM components within 4GB VRAM constraints using maximum and minimal setups.
Video
KGP Talkie
2026-09-16
Do You Know How GLM 5 3 and GLM 5 3 Flash Works? DSA, KDA and KV Cache Explained
Compares GLM 5.3 and GLM 5.3 Flash architectures via config rebuilds, detailing DSA/KDA layer stacking, lightning indexer usage, MoE routing, MTP draft layers, and KV cache calculations.
Video
Luke's Dev Lab
2026-09-16
Qwen 3.8 Max tested with ClinePass
Evaluates Qwen 3.8 Max via ClinePass using performance, memory, agency, OpenAI Human Eval, Kanban, Sand Physics, Dungeon Crawler, Blender and Godot tests within the Cline agent.