BAAI / feature-extraction updated 2 years ago

bge-small-en-v1.5

If you are looking for a model that supports more languages, longer texts, and other retrieval methods, you can try using bge-m3.

Params

30 M

Context

512

Downloads 30d

72.4 M

Likes

525

Commercial use: allowed mit Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

72.9 M69.0 M

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	f32	0.1 GB	0.7 GB	✅ Runs comfortably

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

Specifications

Architecture: BertModel
Parameters: 30 M
Tensor type: F32
Context length: 512
Vocabulary: 30,522
Layers / heads: 12 / 12
Licence: mit
First seen on the Hub: 2023-09-12
Training datasets: undisclosed
MTEB ArguAna (reported): 72.191
MTEB BIOSSES (reported): 83.892272988134
MTEB ArxivClusteringP2P (reported): 47.402770198164
MTEB ArxivClusteringS2S (reported): 40.015454369742
MTEB BiorxivClusteringP2P (reported): 38.467181385006
MTEB BiorxivClusteringS2S (reported): 34.719496037339
MTEB AskUbuntuDupQuestions (reported): 74.421690190388
MTEB Banking77Classification (reported): 85.714935664664
MTEB CQADupstackAndroidRetrieval (reported): 29.587
MTEB AmazonPolarityClassification (reported): 92.738840013315
MTEB AmazonReviewsClassification (en) (reported): 46.559367867279
MTEB AmazonCounterfactualClassification (en) (reported): 68.091494561709
Added to our catalog: 2026-07-28

Family

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

Derivative bge-small-en-v1.5 Xenova 441 K

Compare with

bge-base-en-v1.5 10.0 M bge-small-zh-v1.5 4.8 M granite-embedding-small-english-r2 3.8 M all-MiniLM-L6-v2 2.9 M bge-base-en-v1.5 2.2 M bge-large-zh-v1.5 1.8 M

Appears in

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