Agents & MCPMCP Inspector: How to Use It, CLI Mode and Fixes (v2 Tested)
Run MCP Inspector with one npx command, script it with --cli, and fix the old proxy-token errors. Tested on v2.9.0 against a Python FastMCP server.
Developer guides for local LLMs and AI agent tooling
Run local models, debug MCP servers, and estimate memory and cost, with the commands, versions and test output to reproduce each result. Written by Mian Uzman Munib.
Agents & MCPLatest guides
Local LLMsOllama Alternatives: 9 Local LLM Runtimes Compared (2026)
Nine runtimes that replace Ollama, from LM Studio and llama.cpp to vLLM and Jan, with license, GPU backends, formats, API and last release checked for each.
Local LLMsvLLM vs Ollama: Which to Run for Local Dev or Serving
Ollama wins for one developer on a laptop; vLLM wins for multi-user GPU serving. Here are labeled benchmarks, OS and format support, and a pick for each setup.
Local LLMsOllama Commands: CLI Cheat Sheet, Env Vars, Update and Uninstall
Every ollama command in one table, run on a real install, plus the environment variables, per-OS update and uninstall steps, and where models are stored.
Local LLMs
Local LLMsllama-server: Flags, API, Web UI and Docker (Tested Guide)
Start llama-server with one command, then set context, parallel slots, KV cache and an API key. Tested on llama.cpp v0.5.0 with every endpoint checked.
Agents & MCP
Agents & MCPMCP Inspector: How to Use It, CLI Mode and Fixes (v2 Tested)
Run MCP Inspector with one npx command, script it with --cli, and fix the old proxy-token errors. Tested on v2.9.0 against a Python FastMCP server.
LLM Ops & Costs
LLM Ops & CostsBest Open Source Vector Databases: 4 Picks by Workload
The best open source vector database depends on your data and operating model. Compare pgvector, Qdrant, Milvus, and Weaviate by workload.
