Lesson 2: Ollama vs vLLM vs llama.cpp vs Strata: Choosing a Local Inference Backend in 20262026-10-21 (Modified: 2026-10-21) — Written by SimeonOnSecurity — 11 min readCompare Ollama, vLLM, llama.cpp, and Strata as the inference backend behind an OpenAI-compatible broker. Load lifecycle, health checks, multi-GPU, and a decision table for 2026.#qwen3.8 #hermes agent Learn more →
2x RTX 5060 Ti vs GB10: Five-Model Local AI Benchmark at 256K Context2026-10-10 (Modified: 2026-10-15) — Written by SimeonOnSecurity — 17 min readA five-model RTX 5060 Ti versus GB10 comparison, plus a same-host Qwen3.8 Flash-Next Strata benchmark through 256K context.Local model performance depends on memory architecture, model choice, and context length.#DGX Spark #qwen3.8 #strata #local llm #local inference Learn more →
Claude Cowork and Claude Code With Local Models: 2026 Setup Guide2026-09-22 (Modified: 2026-10-11) — Written by SimeonOnSecurity — 12 min readConfigure Ollama with Claude Desktop and Claude Code through an Anthropic-compatible API. Review model choices and cloud-only features for 2026.#claude code #qwen3.8 #llama.cpp Learn more →
Local AI in 2026: A 27B Model Beats Sonnet 4.6, and the Hardware Window Is Closing2026-09-22 (Modified: 2026-10-11) — Written by SimeonOnSecurity — 16 min readPlan a 2026 local AI rig around model size, VRAM, memory bandwidth, power, cost, and upgrade options.#local ai #qwen3.8 #Qwen3.8 27B #ollama #llama.cpp Learn more →
