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.
#172 of 200 in Embeddings.
On MTEB Indic v1 · Retrieval · multilingual (14 languages), STS Multilingual Mpnet Base v2 scores 52.3 vs 55.4 — lower.
Released 2024-06-07. 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.
STS Multilingual Mpnet Base v2 is an open-weights embeddings & retrieval model from Gameselo, released Jun 2024.
Qwen 3 Embedding 8B (better), Qwen Indic v1 (open weights).
BetterQwen 3 Embedding 8BunstatedOpen+81%—$0.01Memory (4-bit): 0.2 GB → 4.5 GBLOpenQwen Indic v1+82%——Memory (4-bit): 0.2 GB → 4.5 GBiBetterMultilingual E5 Large InstructunstatedOpen+62%—$0.01Memory (4-bit): 0.2 GB → 0.3 GBiBetterMultilingual E5 LargeunstatedOpen+58%—$0.01Memory (4-bit): 0.2 GB → 0.3 GB