How much VRAM does gte-small need?

Q4_K_M is exactly 29,203,744 bytes (0.03 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

thenlper · embedding

gte-small

thenlper/gte-small

gte-small at Q4_K_M is exactly 29,203,744 bytes (0.03 GiB / 0.03 GB) — an effective 7.003 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)From the file· KV per layer

Parameters

33M

Architecture

bert

12 layers

Context

512

native (config.json)

License

mit

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q3_K_S	0.02 GiB	25,250,080	6.055	—	ChristianAzinn
Q2_K	0.02 GiB	25,250,080	6.055	—	ChristianAzinn
Q4_0	0.02 GiB	26,190,112	6.280	—	ChristianAzinn
Q3_K_M	0.02 GiB	26,724,640	6.409	—	ChristianAzinn
Q3_K_L	0.03 GiB	27,738,400	6.652	—	ChristianAzinn
Q4_K_S	0.03 GiB	28,217,632	6.767	—	ChristianAzinn
Q5_0	0.03 GiB	28,844,320	6.917	—	ChristianAzinn
Q4_K_M	0.03 GiB	29,203,744	7.003	—	ChristianAzinn
Q5_K_S	0.03 GiB	29,729,056	7.129	—	ChristianAzinn
Q5_K_M	0.03 GiB	30,475,552	7.308	—	ChristianAzinn
Q6_K	0.03 GiB	35,092,768	8.415	—	ChristianAzinn
Q8_0	0.03 GiB	36,806,944	8.826	—	ChristianAzinn

No KV cache

architectural, not a gap in our data

This architecture allocates no KV cache. Encoder and embedding models process their input in one pass rather than generating token by token, so there is nothing to carry forward between steps and memory does not grow with context. Its footprint is the weights plus a working buffer, and that is the whole story.

Compare with

same modality, comparable size

vs nomic-embed-text-v1.5 vs all-MiniLM-L6-v2 vs mxbai-embed-xsmall-v1 vs jina-reranker-v1-tiny-en vs nomic-embed-text-v1 All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 0.02 GiB. The real file is 0.03 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

384

Vocab

30,522

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does gte-small need?: Q4_K_M is exactly 29,203,744 bytes (0.03 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of gte-small 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.