Home / Models / LFM2.5-24B-A2B

How to Run LFM2.5-24B-A2B Locally — Setup Guide (October 2026)

LFM2.5-24B-A2B is Liquid AI's mixture-of-experts model: 24B total parameters with about 2.45B active per token, so it generates quickly for its size. Liquid publishes the GGUFs itself, and the whole range from Q4_K_M up fits a 16 GB card or two.

Numbers you can trust

We don't print guessed speeds. Every measured number on this page came from TurboLLM's own auto-tuner running LFM2.5-24B-A2B on real hardware — labeled as such. When you run it yourself, TurboLLM benchmarks it on your exact GPU and shows a VRAM-fit verdict before you load, and the real measured tokens/sec once it's running.

Does it fit your GPU?

24B total, about 2.45B active parameters. 40 layers, of which 30 are short-convolution layers that carry no growing KV cache and 10 are full-attention layers (8 KV heads), with 64 experts and 4 routed per token and a 131,072-token context. That is the same conv-plus-attention family as the dense LFM2.5-2.6B, scaled up as a MoE.

LFM2.5-24B-A2B
Liquid AI · GGUF
MoE · ~2.45B active · short-conv + attention hybrid
VRAMFit
16 GBQ4_K_M (14.53 GB) in VRAM with a modest context · 131K maximum context
24 GBQ5_K_M (17.03 GB) or Q6_K (19.69 GB) fully in VRAM · 131K context
32 GBQ8_0 (25.49 GB) fully in VRAM · 131K context
LiquidAI/LFM2.5-24B-A2B-GGUF
View on Hugging Face →

How to run it

  1. Install TurboLLM

    npx turbollm — no install step, works on Windows, macOS, and Linux. It detects your GPU and auto-provisions a matching llama-server build (no CUDA toolkit, no Python, no compiler).

  2. Get the model

    Paste LiquidAI/LFM2.5-24B-A2B-GGUF into TurboLLM's in-app Hugging Face search — it lists every quant with a VRAM-fit verdict against your real free VRAM, and downloads are resumable and SHA-256 verified.

  3. Let auto-fit pick the setup

    Flip the auto-fit toggle and TurboLLM chooses the GPU/CPU layer split (and, for a Mixture-of-Experts model, the expert-offload split) for your card — the same manual flag-hunting every other guide walks you through by hand.

  4. Load it

    TurboLLM benchmarks the model on your exact GPU as it loads and shows the real measured tokens/sec — not a number copied from someone else's card.

  5. Use it

    Chat in the built-in UI, or point any OpenAI- or Anthropic-compatible tool (including Claude Code) at TurboLLM's local API — see the API docs.

One honest caveat

Only 10 of 40 layers keep a growing KV cache, so context is cheap for the size, but the weights still need to fit: on a 16 GB card the Q4_K_M leaves little headroom beyond a short context. No measured speeds are published on this page yet.

FAQ

Will LFM2.5-24B-A2B run on my GPU?

Check the VRAM table above for the tier closest to your card. If you're not sure which tier you're in, or want picks tailored to your exact hardware, use What can I run? — enter your VRAM (or Apple unified memory) and it suggests the model, quant, and offload split that fits, the same logic TurboLLM's auto-fit runs when you load a model.

Which quant should I use for LFM2.5-24B-A2B?

Pick the highest quant that fits your VRAM with room left over for the KV cache and compute buffer — not just the model weights. As a starting point, Q4_K_M is usually the sweet spot (small quality loss for a large size drop); step up to Q5_K_M/Q6_K/Q8_0 if you have headroom, or down to Q3_K_M/an IQ quant if you don't. TurboLLM shows a VRAM-fit verdict for every quant against your card's real free VRAM before you download — see Quantization explained for what the letters and numbers actually mean, or llama.cpp flags that actually matter for the other settings worth knowing.

Why does LFM2.5-24B-A2B suddenly get slow or stutter?

The usual cause is silent VRAM spill: once a model's weights plus its KV cache stop fitting in VRAM, the driver quietly offloads the overflow to system RAM over PCIe, and that path runs 5–10× slower with no error message — just a sudden slowdown. TurboLLM's auto-fit sizes the quant and GPU/CPU split against your card's real free VRAM (including KV-cache growth at your chosen context) specifically to avoid this, and shows a VRAM-fit verdict before you load rather than after it's already crawling. Full mechanics in VRAM spill explained.

Run it now

$ npx turbollm

One command detects your GPU, provisions the right engine, and opens the UI. New here? Start with Install & first run and Quantization explained. On different hardware, see RTX 4070 TI SUPER, RTX 5070 Ti or RTX 3090. Also see how to run it or how to run it.