WUMBOLABS / PROJECTS PROJECT DIRECTORY

/projects/

Projects

WumboLabs publicly builds three projects. Evaluation evidence, long-form reports, and the testing method have dedicated canonical surfaces.

  • Evaluations — structured model evidence and current testing state.
  • Reports — long-form reports, benchmarks, experiments, fit tests, and historical baselines.
  • Methodology — WELP, review rules, claim boundaries, and publication process.
FLAGSHIP PROJECT

LLMGauge

FORMAL RELEASE · v0.78 · PyPI 0.78.0

Local-first CLI for practical LLM evaluation on real consumer hardware.

MAIN FOCUS · ACTIVE DEVELOPMENT

wumbOS

PUBLIC BASELINE · wumbOS v0.2.0-alpha.1 · active development

A developing Linux desktop environment centered on Hyprland, Quickshell, and local-first Agent integration.

Monolith

TECHNICAL PUBLIC ALPHA · v0.11.12

Local-first AI workbench for testing, comparing, and evaluating local LLMs on real hardware.