$0.01 per 1M tokens.
There is no free API; the cheapest paid price is $0.01 per 1M input tokens. The weights are open, so you can run it yourself.
#57 of 200 in Embeddings.
On ChemTEB · Mean over task types · multilingual (13 languages), Bge Large En v1.5 scores 50.7 vs 50.9 — about the same.
Through the BAAI API.
Released 2023-09-12. It is the current version in its line.
Yes, the weights are open (MIT licence).
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Bge Large En v1.5 is an open-weights embeddings & retrieval model from BAAI, released Sep 2023.
Bge Base En v1.5 (cheaper), Cohere Embed English v3 (better), F2LLM v2 8B (open weights).
BetterCohere Embed English v3+6%+900%$0.10Languages: not verifiedcOpenF2LLM v2 8Bunstated+17%——Memory (4-bit): 0.2 GB → 4.5 GBsCheaperAll MiniLM L12 v2unstatedOpen-2%-50%$0.005Max input: 1K tokens → 0K tokenssCheaperAll MiniLM L6 v2unstatedOpen-2%-50%$0.005Max input: 1K tokens → 0K tokens
Text Embedding 3 Largeunstated+12%+1,200%$0.13Languages: not verifiedcF2LLM v2 8BunstatedOpen+17%——Memory (4-bit): 0.2 GB → 4.5 GBcF2LLM v2 14BunstatedOpen+17%——Memory (4-bit): 0.2 GB → 8.4 GBcF2LLM v2 4BunstatedOpen+12%——Memory (4-bit): 0.2 GB → 2.4 GB