Open developer tools

Small, inspectable Python utilities accompanying our technical guides. Download the source, review it, and run it in your own environment. Every tool has explicit assumptions and boundaries. These are not claims of physical GPU testing or independent performance benchmarks.

Before use: inspect any downloaded script and its dependencies. Some tools execute local synthetic tests or create temporary files; others only parse input. The new KV planner models a declared memory budget, not measured serving performance. Python sources are supplied as examples without warranties.

GPU and local LLM serving

Inspect memory assumptions and debug runtime signals without running GPU workloads.

Linux containers and MCP

Bounded diagnostics and synthetic checks for common engineering failures.

Offline research comparisons

Compare previously collected evidence, not fabricated benchmarks.

Found an outdated API, incorrect assumption, or broken tool? Report it here. For testing methodology and AI-assisted editorial review, see our standards.