Models Pie – LLMs ranked by fast / cheap / good tradeoffs
Pie chart trade-off weights beat static leaderboards for model selection.

Opinionated LLM rankings with actual GPU memory math, not just benchmark scores.
Developers self-hosting or deploying open-weight models
LMArena · Artificial Analysis Intelligence Index · Hugging Face Open LLM Leaderboard
Pie chart trade-off weights beat static leaderboards for model selection.
Continual learning pipeline that fine-tunes weights from text feedback, real distributed execution options.
Duplicating transformer layers boosts benchmark scores without a single step of training.
3.9s cold starts vs 45s+ for quantized models—real infra pain solved tangibly.
Direct weight editing for persistent memory—MEMIT meets LoRA consolidation with null-space math.
SQLite-based LLM inference hitting 210MB RSS beats OS paging with deterministic memory control.