Find the best local LLM for your hardware, ranked by benchmarks

v0.5.10 Bugfix

This release fixes issues for SREs watching stability and regressions.

Published 1mo LLM Frameworks

✓ No known CVEs patched

✓ No known CVEs patched in this version

Topics

ai apple-silicon benchmarks cli gguf gpu

+7 more

huggingface inference llm local-llm ollama python vram

Summary

AI summary

Fixed ranking logic so strong partial‑offload candidates are no longer buried and penalties for light and MoE partial offloads are adjusted.

Type	Severity	Summary	CVE
Bugfix
Bugfix	Medium	Fixes double-counting of GPU fit in final sort for strong partial-offload candidates. Fixes double-counting of GPU fit in final sort for strong partial-offload candidates. Source: llm_adapter@2026-06-11 Confidence: high	—
Bugfix	Medium	Reduces aggressive penalty for light partial offload while preserving strong discount for heavy dense offload. Reduces aggressive penalty for light partial offload while preserving strong discount for heavy dense offload. Source: llm_adapter@2026-06-11 Confidence: high	—
Bugfix	Medium	Applies milder penalty for MoE partial-offload when active working set can stay on GPU. Applies milder penalty for MoE partial-offload when active working set can stay on GPU. Source: llm_adapter@2026-06-11 Confidence: high	—

Full changelog

Strong partial-offload candidates no longer get buried under weaker full-GPU models because the final sort no longer counts GPU fit twice.
Light partial offload is penalized less aggressively, while heavy dense offload still gets a strong discount.
MoE partial-offload scoring now gives a milder penalty when the active working set can plausibly stay on GPU.

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track Find the best local LLM for your hardware, ranked by benchmarks

Get notified when new releases ship.

About Find the best local LLM for your hardware, ranked by benchmarks