What are the hardware requirements for Diffusion Lfm2.5 24B A2b?

Diffusion Lfm2.5 24B A2b requires approximately 96.1 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Diffusion Lfm2.5 24B A2b and how large is it?

Diffusion Lfm2.5 24B A2b is developed by summerAIO, a model with 24b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Diffusion Lfm2.5 24B A2b?

Diffusion Lfm2.5 24B A2b is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Diffusion Lfm2.5 24B A2b by summerAIO — VRAM 96.1GB, 128K context

Name: Diffusion Lfm2.5 24B A2b
Author: summerAIO

Diffusion Lfm2.5 24B A2b is an open-source language model by summerAIO. Features: 24b LLM, VRAM: 96.1GB, Context: 128K, LLM Explorer Score: 0.37.

Code-generation Conversational Custom code Diffusion Endpoints compatible Experimental Lfm2 Lfm2 moe Lora Lora-merged Moe Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/summerAIO/diffusion-lfm2.5-24b-a2b

Diffusion Lfm2.5 24B A2b Parameters and Internals

LLM Name	Diffusion Lfm2.5 24B A2b
Repository 🤗	https://huggingface.co/summerAIO/diffusion-lfm2.5-24b-a2b
Model Size	24b
Required VRAM	96.1 GB
Updated	2026-11-11
Maintainer	summerAIO
Model Type	lfm2_moe
Model Files	0.0 GB 4.3 GB: 1-of-12 3.8 GB: 1-of-13 4.3 GB: 2-of-12 3.7 GB: 2-of-13 4.3 GB: 3-of-12 3.7 GB: 3-of-13 4.3 GB: 4-of-12 3.8 GB: 4-of-13 4.3 GB: 5-of-12 3.7 GB: 5-of-13 4.3 GB: 6-of-12 3.7 GB: 6-of-13 4.3 GB: 7-of-12 3.7 GB: 7-of-13 4.3 GB: 8-of-12 3.8 GB: 8-of-13 4.3 GB: 9-of-12 3.7 GB: 9-of-13 4.3 GB: 10-of-12 3.7 GB: 10-of-13 4.3 GB: 11-of-12 3.7 GB: 11-of-13 0.7 GB: 12-of-12 3.8 GB: 12-of-13 3.3 GB: 13-of-13
Model Architecture	Lfm2MoeForCausalLM
Context Length	131072
Model Max Length	131072
Transformers Version	5.22.2
Tokenizer Class	TokenizersBackend
Padding Token	<\|pad\|>
Vocabulary Size	128000
LoRA Model	Yes

Rank the Diffusion Lfm2.5 24B A2b capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Diffusion Lfm2.5 24B A2b

Best Alternatives	Context / RAM	Downloads	Likes
LFM2.5 24B A2B	128K / 48.3 GB	9356	424
LFM2.5 24B A2B FP8	128K / 24.6 GB	1691	22
LFM2.5 24B A2B	128K / 48.3 GB	2243	8
LFM2.5 24B A2B Base	128K / 48.3 GB	362	35
LFM2.5 24B A2B ONNX	128K / GB	546	20
LFM2.5 24B A2B MLX Bf16	128K / 48.2 GB	1073	6
LFM2.5 24B A2B DERISKED BF16	128K / 48.3 GB	458	3
LFM2.5 24B A2B Preview Bf16	128K / 48.2 GB	596	1
LFM2.5 24B A2B NVFP4 W4A16	128K / 14.9 GB	239	1
LFM2.5 24B A2B NVFP4A16	128K / 0.3 GB	230	1

Note: green Score (e.g. "73.2") means that the model is better than summerAIO/diffusion-lfm2.5-24b-a2b.