How much VRAM does LFM2.5-Audio-1.5B need?

Q4_0 is exactly 855,282,912 bytes (0.80 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

LiquidAI · text

LFM2.5-Audio-1.5B

LiquidAI/LFM2.5-Audio-1.5B

LFM2.5-Audio-1.5B at Q4_0 is exactly 855,282,912 bytes (0.80 GiB / 0.86 GB) — an effective 4.654 bits per weight, not the nominal 4.

From the file· summed from 3 file(s)From the file· KV per layer

Parameters

1.5B

Architecture

lfm2

null layers

Context

—

native (config.json)

License

other

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_03 shards	0.80 GiB	855,282,912	4.654	—	Paulescu
Q4_03 shards	0.80 GiB	855,283,552	4.654	—	LiquidAI
Q8_03 shards	1.42 GiB	1,528,952,544	8.319	—	Paulescu
Q8_03 shards	1.42 GiB	1,528,953,184	8.319	—	LiquidAI
F163 shards	2.68 GiB	2,873,183,904	15.633	—	Paulescu
F163 shards	2.68 GiB	2,873,184,544	15.633	—	LiquidAI

Compare with

same modality, comparable size

vs Qwen3-4B vs HyperCLOVAX-SEED-Text-Instruct-1.5B vs Llama-3.2-1B-Instruct vs Qwen3-VL-8B-Instruct-abliterated-v1 vs ced-base vs gemma-3-1b-it All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_0 at roughly 0.77 GiB. The real file is 0.80 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json

Layers

—

Attention heads

—

KV heads

—

Head dim

—

Hidden size

—

Vocab

—

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does LFM2.5-Audio-1.5B need?: Q4_0 is exactly 855,282,912 bytes (0.80 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of LFM2.5-Audio-1.5B should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.