Local AI image/video user deletes 3.2TB of old models, keeps almost nothing
legarth · reddit · 2026-09-07
A ComfyUI user, running out of SSD space, deleted 3.2TB of models accumulated over four years, keeping only Ideogram 4, LTX 2.5, Z-image, plus tools like SAM and VibeVoice. With Krea 2 and MiniMax H3 available, they found they never went back to anything between SD1.4 and LTX 2.3 — the 'nostalgia generation' they expected never happened. Even the SDXL folder is gone. The thread debates whether keeping old model checkpoints is worth it.
More from Multimodal
- FLUX.3 Image Klein tested across 15 complex prompts: strong on text, weak on distant faces — SkirtSwimming8950 · 2026-10-23
- Two Midjourney sref codes for grainy pastel fashion illustration style — sergeantsref · 2026-10-23
- Three free ways to try MOSS-Image: API playground, HF Space, ComfyUI — songguoxiansen · 2026-10-23
- Open-source MOSS-Image tops 1M ComfyUI downloads, 5B model rivals closed rivals — songguoxiansen · 2026-10-23
- Testing light effects in MiniMax video generation — the guinea pig was fine — seppe0815 · 2026-10-23
- GAFU Uses Generation-Side Representations to Boost Fine-Grained Visual Understanding in Unified Multimodal Models — Tengju Ru · 2026-10-23