multilingual-e5-base vs all-MiniLM-L6-v2-GGUF
Specs, VRAM requirements and download trends — updated 19 September 2026.
intfloat · A · Open weights
Params
280 M
Context
512
30d
4.9 M
mit · commercial OK
leliuga · B · Open weights
Params
20 M
Context
512
30d
396 K
apache-2.0 · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | multilingual-e5-base | all-MiniLM-L6-v2-GGUF |
|---|---|---|
| Parameters | 280 M | 20 M |
| Architecture | XLMRobertaModel | BertModel |
| Context length | 512 | 512 |
| Licence | mit | apache-2.0 |
| Commercial use | Allowed | Allowed |
| Languages | 94 | 1 |
| Downloads 30d | 4,907,595 | 396,202 |
| Downloads all time | 75.0 M | 500 K |
| Quantizations on the Hub | SAFETENSORS | GGUF |
| Gated | No | No |
| First seen on the Hub | 2023-05-19 | 2024-03-15 |
Download trend
Daily snapshots, last 90 days (13 Aug – 10 Nov)
7.6 M396 K
multilingual-e5-base
all-MiniLM-L6-v2-GGUF
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
all-MiniLM-L6-v2-GGUF · Q4_0
0.5 GB / 24 GB
multilingual-e5-base · fp16
1.8 GB / 24 GB