bge-reranker-base vs roberta_toxicity_classifier

Specs, VRAM requirements and download trends — updated 24 September 2026.

BAAI · A · Open weights
Params
280 M
Context
512
30d
2.4 M
mit · commercial OK
s-nlp · B · Open weights
Params
—
Context
512
30d
231 K
openrail++ licence · conditional

Specification comparison

Differences are highlighted; identical values are muted.

Specification bge-reranker-base roberta_toxicity_classifier
Parameters 280 M —
Architecture XLMRobertaForSequenceClassification RobertaForSequenceClassification
Context length 512 512
Licence mit openrail++
Commercial use Allowed Conditional
Languages 2 1
Downloads 30d 2,364,917 230,823
Downloads all time 54.6 M 5.0 M
Quantizations on the Hub SAFETENSORS —
Gated No No
First seen on the Hub 2023-09-11 2022-03-02

Download trend

Daily snapshots, last 90 days (13 Aug – 10 Nov)

4.2 M222 K
bge-reranker-base roberta_toxicity_classifier

VRAM side by side

On RTX 4090 · 24 GB · 8K context unless noted

bge-reranker-base · fp16 1.8 GB / 24 GB
✅ Runs comfortably

Adjacent comparisons