Benchmark Lab

Honest comparisons. Real numbers, stated methodology, no vendor copy. These are our measurements — run your own.

All internal testing. Not sourced from vendors. See methodology on each page.

Benchmark Lab
Active Benchmarks
LLM Code Completion: Speed vs Quality
Four models, 120 prompts, real latency numbers. How does token generation speed correlate with output quality for developer use cases?
May 2026 · 4 models · @rkovacs_ml
Terminal Emulator Shootout 2026
Ghostty vs Wezterm vs Alacritty vs Kitty. Startup time, memory at rest, font rendering, and practical notes from daily use.
Apr 2026 · 4 emulators · @mschubert_dev
Git Hosting Platform Comparison
GitHub vs GitLab vs Forgejo vs Sourcehut. Feature matrix, not a ranking. What you get, what you trade off.
Mar 2026 · 4 platforms · @mschubert_dev