OPEN LINK ↗
// 75d agoPRODUCT UPDATE
Harbor 0.4.4 unifies local model management
Harbor v0.4.4 adds cross-backend model lifecycle commands so developers can list, pull, and remove models for Ollama, Hugging Face, and llama.cpp from one CLI. The update reduces day-to-day friction for local LLM setups that mix multiple runtimes and caches.
// ANALYSIS
This is a practical quality-of-life release that solves an operational pain every serious local LLM user eventually hits.
- –`harbor models ls` gives one inventory view across three model sources, including JSON output for scripting.
- –`harbor models pull` intelligently routes model specs between Hugging Face and Ollama, which simplifies fetch logic in automation.
- –`harbor models rm` removes matched models across caches, helping teams avoid stale artifacts and reclaim disk without manual cleanup.
// TAGS
harborllmclidevtoolinferenceopen-sourceself-hosted
DISCOVERED
75d ago
2026-03-14
PUBLISHED
75d ago
2026-03-13
RELEVANCE
8/ 10
AUTHOR
Everlier