Stop relying on parameter count guesses. Learn how to calculate the four real memory consumers: weights, KV cache, activation, and runtime overhead, so you can avoid out-of-memory crashes.
Learn how to find your actual usable free RAM and VRAM on Windows, macOS, and Linux—or use the whichllmmodel CLI to inspect it in seconds, to prevent out-of-memory crashes.
How modern MoE architectures like Qwen 3.6 35B A3B and Gemma 4 26B differ from dense models, what lives inside each layer, and why --cpu-moe saves speed for MoE models, when compared to standard -ngl.
Understand the practical differences between SWE-bench Pro, SWE-bench Verified, LiveCodeBench, and HumanEval — and choose the exact model tier needed for your developer stack.