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.
#139 of 200 in Embeddings.
On MTEB multilingual v2 · Reranking · multilingual (22 languages), Sentence Croissant Alpha v0.3 scores 51.5 vs 51.6 — about the same.
Released 2024-04-26. 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.
Sentence Croissant Alpha v0.3 is an open-weights embeddings & retrieval model from manu, released Apr 2024.
EmbeddingGemma 300M (better), F2LLM v2 8B (open weights).
BetterEmbeddingGemma 300MunstatedOpen+23%—$0.002cOpenF2LLM v2 8Bunstated+37%——Memory (4-bit): 0.8 GB → 4.5 GB
BetterQwen 3 Embedding 8BunstatedOpen+28%—$0.01Memory (4-bit): 0.8 GB → 4.5 GB
BetterBge M3unstatedOpen+22%—$0.01iBetterMultilingual E5 Large InstructunstatedOpen+22%—$0.01
Llama Embed Nemotron 8BOpen+32%——Memory (4-bit): 0.8 GB → 4.5 GBOOcten Embedding 8BOpen+31%——Memory (4-bit): 0.8 GB → 4.5 GB