Lfm 2 24B · Ollama

lfm2

LFM2 is a family of hybrid models designed for on-device deployment. LFM2-24B-A2B is the largest model in the family, scaling the architecture to 24 billion parameters while keeping inference efficient.

tools 24b

1.1M

richardyoung/lfm2.5-24b-a2b

LiquidAI LFM2.5-24B-A2B: 24B reasoning MoE with only 2.45B active. Fast agents and tool use, 16 languages.

38

dlasher/LFM2.5-24B-A2B-GGUF

pulled as BF16, requant to IQ4_NL, with combined_en_huge as dataset (English + Coding focus)

15

dlasher/LFM2.5-24B-A2B-ENG

pulled as BF16, vocab slice (combined_en_huge) requant to IQ4_NL (combined_en_huge) -- (English + Coding focus) - ~20% faster than Q4_K_M

13

supergoatscriptguy/mythos-sec

Local security/pentest assistant. Now on LFM2 24B MoE, 30B-class quality at 8B-class speed. Tuned for CTF, bug bounty, pentest methodology, vulnerability analysis. Tool-calling ready, 32k ctx. Concise, direct, no disclaimers. Smaller 8b variant available.

visiontoolsthinking 8b 24b

4,539

supergoatscriptguy/mythos-code

Fast local coding agent on LFM2 24B MoE. 30B-class quality at 8B-class speed. Tool-calling ready, 32k context. Tuned for Aider, Cline, Continue, Cursor, Codex, OpenCode, and Claude Code workflows.

tools 24b

538

Maternion/lfm2

LFM2-8B-A1B is the best on-device MoE in terms of both quality (comparable to 3-4B dense models) and speed (faster than Qwen3-1.7B). Updated LFM2.5-8B-A1B is now available!

tools 8b

2,182

jewelzufo/LFM2.5-1.2B-Distilled-Claude-4.6

Source: https://huggingface.co/FlameF0X/LFM2.5-1.2B-Distilled-Claude-4.6-GGUF?local-app=ollama

297

oamazonasgabriel/lfm2.5-2.6b

LFM2.5-2.6B by Liquid AI — deploy agents everywhere. A 2.6B dense reasoning model (Q4_K_M, ~1.7 GB) with native tool calling, 128K context, and 16-language support. Runs on any 8 GB GPU.

1,905

Models