Liquid AI - Mixture of experts - verified
Run Liquid AI LFM2.5-24B-A2B locally
Liquid AI LFM2.5-24B-A2B is a 24B mixture-of-experts (2.5B active) model from Liquid AI with a 131K-token context window, released under LFM Open License v1.0 (lfm1.0) (commercial use is allowed with conditions set out in the license); the smallest listed download is 13.6 GB, the 4bit MLX build from LiquidAI/LFM2.5-24B-A2B-MLX-4bit.
Official weights: LiquidAI/LFM2.5-24B-A2B
- Total params
- 24B
- Active params
- 2.5B
- Context
- 131K
- Smallest download
- 13.6 GB
Fits in
Smallest download at most 85% of memoryDownloads
The official weights and every quantization listed for this model, with file sizes from Hugging Face.
| Download | Format | Size | Repo |
|---|---|---|---|
Official weights | BF16 | 47.9 GB | LiquidAI/LFM2.5-24B-A2B |
Q4_K_M | GGUF | 14.5 GB | LiquidAI/LFM2.5-24B-A2B-GGUF |
Q8_0 | GGUF | 25.5 GB | LiquidAI/LFM2.5-24B-A2B-GGUF |
4bitSmallest listed | MLX | 13.6 GB | LiquidAI/LFM2.5-24B-A2B-MLX-4bit |
GB is decimal: 1 GB = 10^9 bytes. Add memory for the context and the operating system on top of the file size.
Run it
Copied verbatim from the page linked under each command. Server commands from model cards often assume a multi-GPU node; check the flags before you run them.
Not sure which runtime to use? Compare local LLM runtimes by platform, API and features.
License
- License
- LFM Open License v1.0 (lfm1.0)
other - Commercial use
- Commercial use: ConditionsCommercial use is allowed with conditions set out in the license.
- Gated download
- No. The files download directly.
Worth knowing
Pure reasoning model that always thinks before answering (the chat template opens a think tag) and needs transformers 5.0.0 or newer.
Released (Hugging Face repo creation date).
FAQ
How big is the Liquid AI LFM2.5-24B-A2B download?
The official BF16 weights in LiquidAI/LFM2.5-24B-A2B are 47.9 GB. The smallest listed download is 13.6 GB, the 4bit MLX build from LiquidAI/LFM2.5-24B-A2B-MLX-4bit. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.
Can I use Liquid AI LFM2.5-24B-A2B commercially?
Liquid AI LFM2.5-24B-A2B is released under LFM Open License v1.0 (lfm1.0), and commercial use is allowed with conditions set out in the license. The license text is at https://huggingface.co/LiquidAI/LFM2.5-24B-A2B/blob/main/LICENSE.
How much memory do I need for Liquid AI LFM2.5-24B-A2B?
By this directory's rule of thumb it fits in 16 GB and up: the smallest listed download, 13.6 GB, is at most 85% of 16 GB, which leaves room for the context and the operating system.
How do I run Liquid AI LFM2.5-24B-A2B with Ollama?
Run "ollama run lfm2.5:24b", as shown at https://ollama.com/library/lfm2.5:24b. The Ollama library lists the lfm2.5:24b tag at 15GB.
What is the context length of Liquid AI LFM2.5-24B-A2B?
131,072 tokens. A longer context needs more memory for the KV cache, so you can run with a shorter one when memory is tight.
Sources
Last verified from the pages below.
- HF API (params, license, createdAt) - huggingface.co
- HF file tree - huggingface.co
- Model card - huggingface.co
- Ollama tags - ollama.com
- Liquid GGUF tree - huggingface.co
- Liquid MLX 4-bit tree - huggingface.co
Newsletter
New open weights, every week
One email per week. New model releases, open-weight drops and what they mean for what you build.
Keep exploring
Best local AI models
Our curated picks by hardware tier, with benchmarks and the reasoning.
Model API pricing and specs
First-party prices per 1M tokens when you would rather call an API.
AI benchmark leaderboards
Sourced scores, for when you need to compare quality.
Setup guides
Point your coding tools at the model you picked.