S

Blip Image Captioning Large

Dec 2023
SBlip Image Captioning Large
Compare in
on
self-host onlyprice
1.1MIEB lite
Dec 2023released
768Dimensions
1K tokensMax input

Ranking?

Full ranking →
1EVA02 CLIP B 1692.42CLIP ViT L 14 DataComp.XL S13B B90K86.43Jina Embeddings v5 Omni Small864CLIP ViT BigG 14 Laion2B 39B B160k85.65CLIP ViT H 14 Laion2B S32B B79K83.9
···
31Moco v3 Vit B73.932Blip Image Captioning Large73.733VLM2Vec LoRA72.6

Strong at

C
Embeddings & Retrieval#88 of 89

Alternatives?

ModelScoreCostPriceWhat changes
BetterJina Embeddings v5 Omni SmallunstatedOpen+17%—$0.05Memory (4-bit): 0.3 GB → 1 GBQOpenEVA02 CLIP B 16unstated+25%——Max input: 1K tokens → 0K tokens+1VBetterVoyage Multimodal 3+12%—$0.12lBetterCLIP ViT L 14 DataComp.XL S13B B90KOpen+17%——Max input: 1K tokens → 0K tokenslBetterCLIP ViT BigG 14 Laion2B 39B B160kOpen+16%——Max input: 1K tokens → 0K tokens+1

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
LicenceBSD-3-Clausefree for commercial use

Benchmarks

in Embeddings & Retrieval
MIEB English · Document Understanding · English0.8#81
MIEB English · Zero-shot Classification · English9.2#39
MIEB English · Visual semantic similarity (English) · multilingual (13 languages)62.2#31
MIEB English · Image Classification · English66.4#39
MIEB English · Compositionality · English27.1#39
MIEB English · Vision Centric QA · English49.1#19
MIEB English · Image Clustering · English73.7#32
MIEB Img · Image Classification · English66.4#39
MIEB Img · Visual semantic similarity (English) · English60.5#30
MIEB Img · Visual semantic similarity (multilingual) · multilingual (13 languages)36.6#29
MIEB Img · Image Clustering · English73.7#32
MIEB lite · Mean over tasks · multilingual (41 languages)31.1#46
MIEB lite · Mean over task types · multilingual (41 languages)31.1#46
MIEB lite · Vision Centric QA · English51#26
MIEB lite · Any-to-any Retrieval · English13#46
MIEB lite · Document Understanding · English1.1#88
MIEB lite · Zero-shot Classification · English3.8#46
MIEB lite · Compositionality · English23.2#47
MIEB lite · Image Classification · English61.8#50
MIEB lite · Any-to-any Multilingual Retrieval · multilingual (40 languages)0.4#47
MIEB lite · Image Clustering · English64.6#41
MIEB lite · Visual semantic similarity (English) · English60.1#41
MIEB lite · Visual semantic similarity (multilingual) · multilingual (13 languages)31.9#35
MIEB multilingual · Zero-shot Classification · English9.2#39
MIEB multilingual · Image Classification · English66.4#39
MIEB multilingual · Document Understanding · English0.8#81
MIEB multilingual · Compositionality · English27.1#39
MIEB multilingual · Vision Centric QA · English49.1#19
MIEB multilingual · Visual semantic similarity (English) · multilingual (13 languages)62.2#31
MIEB multilingual · Visual semantic similarity (multilingual) · multilingual (13 languages)31.9#35
MIEB multilingual · Any-to-any Multilingual Retrieval · multilingual (40 languages)0.3#38
MIEB multilingual · Image Clustering · English73.7#32

FAQ

How much does Blip Image Captioning Large cost?›

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

Is Blip Image Captioning Large free to use?›

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

How good is Blip Image Captioning Large, and what is it best at?›

#88 of 89 in Multimodal.

Is Blip Image Captioning Large better than Moco v3 Vit B?›

On MIEB multilingual · Image Clustering · English, Blip Image Captioning Large scores 73.7 vs 73.9 — about the same.

When was Blip Image Captioning Large released? Is there a newer version?›

Released 2023-12-07. It is the current version in its line.

Is Blip Image Captioning Large open source?›

Yes, the weights are open (bsd-3-clause licence).

Can I run Blip Image Captioning Large locally?›

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

What is Blip Image Captioning Large?›

Blip Image Captioning Large is an open-weights embeddings & retrieval model from Salesforce, released Dec 2023.

What are the best alternatives to Blip Image Captioning Large?›

Jina Embeddings v5 Omni Small (better), EVA02 CLIP B 16 (open weights).