Ollama Library · lfm2moe
Can I run 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.
23.8B parameters32K contexttextLFM Open License v1.02 builds
lfm2 system requirements
Memory needed at 4K context, counting model weights, KV cache, compute buffers and runtime overhead. Download sizes come from the official Ollama registry.
| Build | Quantization | Download | Memory needed |
|---|---|---|---|
| lfm2:24b | Q4_K_M | 13.4 GB | ~19.6 GB(est.) |
| lfm2:24b-q8_0 | Q8_0 | 23.6 GB | ~34.2 GB(est.) |
Figures marked (est.) approximate the KV cache because this model's architecture details are not published. They are indicative rather than exact.
How to run lfm2 locally
- 1.Install Ollama for Windows, macOS or Linux.
- 2.Run this in a terminal:
ollama run lfm2:24b - 3.The weights download on first run, then an interactive prompt opens.
Frequently asked questions about lfm2
- How much VRAM do I need to run lfm2?
- At 4K context, the smallest published build of lfm2 (Q4_K_M) needs roughly 19.6 GB of GPU memory once weights, KV cache, compute buffers and runtime overhead are counted. Bigger quantizations and longer context windows need more. It can also run partly or entirely on the CPU using system RAM, more slowly.
- Can I run lfm2 without a dedicated GPU?
- Yes, but slowly. Without a GPU the model runs on the CPU using system RAM, which usually means a few tokens per second rather than dozens. You would need at least 19.6 GB of free RAM for the smallest build.
- How big is the lfm2 download?
- The smallest published build is 13.4 GB. There are 2 builds in total, the largest being 23.6 GB. Leave some extra free disk space beyond the download itself.
- How do I run lfm2 locally?
- Install Ollama, then run "ollama run lfm2:24b" in a terminal. The weights download on first use and an interactive session opens.
Will lfm2 run on your PC?
Scan your hardware once and get a verdict for this model and every build — free, no account.
Model data from the official Ollama registry · verified 13/09/2026