Fresh Drops · Updated Apr 25, 2026

Newest Open-Source Models

The 12 latest models from the 176-model LocalClaw catalogue, including Ling-2.6-flash and Qwen 3.6, ready to search and run in LM Studio.

Sorted by release date · Most recent first · ⭐ Now includes Ling-2.6-flash + Qwen 3.6

Latest add

Ling-2.6-flash

104B MoE with 7.4B active parameters, MIT license, 262K context and agent/tool-use focus.

Catalogue

176 local LLMs

Cards are hydrated from js/data.js, matching the main recommender catalogue.

Speech catalogue

47 TTS/ASR models

Speech models remain available from the dedicated TTS/ASR page.

Current latest drops

Ling-2.6-flash is now live on LocalClaw

The New page is being hydrated from the 176-model catalogue. If the dynamic cards do not appear yet, the newest entries are still available below.

Apr 2026 · Latest Ling-2.6-flash 80 GB RAM · Q4_K_M est. · 262K context Apr 2026 · Compact Qwen 3.6 (6.7B) 8 GB RAM · Q4_K_M · hybrid thinking Apr 2026 · Flagship Qwen 3.6 (27B) 32 GB RAM · Q4_K_M · reasoning/coding

Want a personalized pick?

Tell us your hardware and we'll find the best model for you in 30 seconds.

Find My Model → Browse All Models