l

CLIP ViT B 16 DataComp.XL S13B B90K

Apr 2023
lCLIP ViT B 16 DataComp.XL S13B B90K
Compare in
on
self-host onlyprice
22.7MIEB lite
Apr 2023released
0K tokensMax input

Ranking?

Full ranking →
1LCO Embedding Omni 7B82.42LCO Embedding Omni 3B79.53E5 Omni 7B77.34EmbeddingGemma 276.85E5 Omni 3B73.5
···
39Clip Vit Base Patch1629.840CLIP ViT B 16 DataComp.XL S13B B90K28.541EVA02 CLIP B 1628.3

Strong at

C
Embeddings & Retrieval#73 of 89

Alternatives?

ModelScoreCostPriceWhat changes
BetterEmbeddingGemma 2unstatedOpen×2.7—$0.02Memory (4-bit): 0.1 GB → 0.2 GBLOpenLCO Embedding Omni 7Bunstated×2.9——Memory (4-bit): 0.1 GB → 5.4 GBBetterJina Embeddings v5 Omni NanounstatedOpen+89%—$0.02Memory (4-bit): 0.1 GB → 0.6 GBBetterJina Embeddings v5 Omni SmallunstatedOpen×2.3—$0.05Memory (4-bit): 0.1 GB → 1 GBBetterJina Clip v1Open+21%—$0.05

Run it yourself

estimate
Fits on✓ 8 GB (phone, laptop) · full quality✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceMITfree for commercial use

Benchmarks

in Embeddings & Retrieval
MIEB English · Mean over tasks · multilingual (13 languages)55.8#17
MIEB English · Mean over task types · multilingual (13 languages)55.8#17
MIEB English · Document Understanding · English27#64
MIEB English · Zero-shot Classification · English62.7#14
MIEB English · Any-to-any Retrieval · English34.8#18
MIEB English · Visual semantic similarity (English) · multilingual (13 languages)68.5#14
MIEB English · Image Classification · English78.6#17
MIEB English · Compositionality · English38.6#27
MIEB English · Vision Centric QA · English54.3#4
MIEB English · Image Clustering · English81.7#14
MIEB Img · Mean over tasks · multilingual (13 languages)59.7#16
MIEB Img · Mean over task types · multilingual (13 languages)59.7#16
MIEB Img · Image Classification · English78.6#17
MIEB Img · Any-to-any Retrieval · English37.2#19
MIEB Img · Visual semantic similarity (English) · English67.1#14
MIEB Img · Visual semantic similarity (multilingual) · multilingual (13 languages)33.8#35
MIEB Img · Image Clustering · English81.7#14
MIEB lite · Mean over tasks · multilingual (41 languages)46.6#29
MIEB lite · Mean over task types · multilingual (41 languages)46.6#29
MIEB lite · Vision Centric QA · English56.9#8
MIEB lite · Any-to-any Retrieval · English28.3#24
MIEB lite · Document Understanding · English22.7#73
MIEB lite · Zero-shot Classification · English61.9#20
MIEB lite · Compositionality · English31.4#33
MIEB lite · Image Classification · English73.2#19
MIEB lite · Any-to-any Multilingual Retrieval · multilingual (40 languages)19.9#32
MIEB lite · Image Clustering · English73.6#22
MIEB lite · Visual semantic similarity (English) · English69.6#20
MIEB lite · Visual semantic similarity (multilingual) · multilingual (13 languages)28.5#40
MIEB multilingual · Mean over tasks · multilingual (41 languages)49.6#23
MIEB multilingual · Mean over task types · multilingual (41 languages)49.6#23
MIEB multilingual · Zero-shot Classification · English62.7#14
MIEB multilingual · Image Classification · English78.6#17
MIEB multilingual · Document Understanding · English27#64
MIEB multilingual · Any-to-any Retrieval · English34.8#18
MIEB multilingual · Compositionality · English38.6#27
MIEB multilingual · Vision Centric QA · English54.3#4
MIEB multilingual · Visual semantic similarity (English) · multilingual (13 languages)68.5#14
MIEB multilingual · Visual semantic similarity (multilingual) · multilingual (13 languages)28.5#40
MIEB multilingual · Any-to-any Multilingual Retrieval · multilingual (40 languages)21.6#23
MIEB multilingual · Image Clustering · English81.7#14

FAQ

How much does CLIP ViT B 16 DataComp.XL S13B B90K cost?›

There is no per-unit API price: self-host only.

Is CLIP ViT B 16 DataComp.XL S13B B90K free to use?›

There is no free API. The weights are open, so you can run it yourself.

How good is CLIP ViT B 16 DataComp.XL S13B B90K, and what is it best at?›

#73 of 89 in Multimodal.

Is CLIP ViT B 16 DataComp.XL S13B B90K better than Clip Vit Base Patch16?›

On MIEB multilingual · Visual semantic similarity (multilingual) · multilingual (13 languages), CLIP ViT B 16 DataComp.XL S13B B90K scores 28.5 vs 29.8 — lower.

When was CLIP ViT B 16 DataComp.XL S13B B90K released? Is there a newer version?›

Released 2023-04-26. It is the current version in its line.

Is CLIP ViT B 16 DataComp.XL S13B B90K open source?›

Yes, the weights are open (MIT licence).

Can I run CLIP ViT B 16 DataComp.XL S13B B90K locally?›

The weights are open and can be self-hosted; hardware needs are not confirmed yet.

What is CLIP ViT B 16 DataComp.XL S13B B90K?›

CLIP ViT B 16 DataComp.XL S13B B90K is an open-weights embeddings & retrieval model from laion, released Apr 2023.

What are the best alternatives to CLIP ViT B 16 DataComp.XL S13B B90K?›

EmbeddingGemma 2 (better), LCO Embedding Omni 7B (open weights).