LiquidAI / text-generation updated 1 day ago

LFM2.5-8B-A1B

Name: LFM2.5-8B-A1B
Author: LiquidAI

LFM2.5 is a new family of hybrid models designed for on-device deployment. It builds on the LFM2 architecture with extended pre-training and reinforcement learning.

Params

8.5 B

Context

128,000

Downloads 30d

172 K

Likes

700

Licence: other Not gated SAFETENSORS 10 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

172 K159 K

Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	bf16	16.9 GB	20.4 GB	⚠️ Tight — reduce context
model.safetensors (bf16, full)	bf16 + 125K ctx	16.9 GB	39.0 GB	❌ Won’t fit

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub

~ · curl · api/v1

$ curl -s https://aimodelscomparison.com/api/v1/models/lfm2-5-8b-a1b
{
  "hf_id": "LiquidAI/LFM2.5-8B-A1B",
  "params_b": 8.47,
  "context_length": 128000,
  "license": { "id": "other", "commercial": "unknown" },
  "downloads_30d": 171520,
  "vram_estimates": [
    { "quant": "bf16", "gb": 20.4 }
  ],
  "updated_at": "2026-08-05T01:00:32Z"
}

est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture: Lfm2MoeForCausalLM
Parameters: 8.5 B
Tensor type: BF16
Context length: 128,000
Vocabulary: 128,000
Layers / heads: 24 / 32
Licence: other
First seen on the Hub: 2026-05-28
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

Base model and the most-downloaded derivatives in the catalog.

Derivative OTel-LLM-8B-A1B-IT farbodtavakkoli 1.4 M

Compare with

Qwen3-8B 16.3 M Qwen2.5-7B-Instruct 12.1 M Qwen3.6-35B-A3B-NVFP4 11.3 M Llama-3.1-8B-Instruct 8.0 M Qwen2.5-7B-Instruct-AWQ 4.6 M granite-4.1-8b 4.2 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB