There is no per-unit API price: self-host only.
There is no free API. The weights are open, so you can run it yourself.
#75 of 89 in Multimodal.
On MIEB English · Zero-shot Classification · English, Clip Vit Base Patch16 scores 56.4 vs 57.6 — lower.
Released 2021-02-26. It is the current version in its line.
Yes, the weights are open.
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Clip Vit Base Patch16 is an open-weights embeddings & retrieval model from OpenAI, released Feb 2021.
Jina Embeddings v5 Omni Small (better), Siglip So400m Patch14 384 (open weights).
BetterJina Embeddings v5 Omni SmallunstatedOpen+12%—$0.05Memory (4-bit): 0.1 GB → 1 GB
OpenSiglip So400m Patch14 384+26%——Memory (4-bit): 0.1 GB → 0.5 GBQBetterEVA02 CLIP B 16unstatedOpen+26%——Memory (4-bit): 0.1 GB → 3 GBlBetterCLIP ViT BigG 14 Laion2B 39B B160kOpen+23%——Memory (4-bit): 0.1 GB → 1.5 GBlBetterCLIP ViT L 14 DataComp.XL S13B B90KOpen+21%——Memory (4-bit): 0.1 GB → 0.3 GB