There is no per-unit API price: API price not public.
There is no free API. The weights are open, so you can run it yourself.
#107 of 200 in Embeddings.
On MTEB Spanish v1 · STS · multilingual (2 languages), Bekko Embedding v1 A25m scores 79.3 vs 79.3 — about the same.
Released 2026-07-28. 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.
Bekko Embedding v1 A25m is an open-weights embeddings & retrieval model from hotchpotch, released Jul 2026.
Jina Embeddings v5 Text Nano (better), Jina Embeddings v5 Text Small (open weights).
BetterJina Embeddings v5 Text NanounstatedOpen+3%—$0.02
BetterGemini Embedding 001unstated+5%—$0.15Max input: 8K tokens → 2K tokensiOpenMultilingual E5 Large Instructunstatedsame—$0.01Max input: 8K tokens → 1K tokens+1
OpenGte Qwen2 7B Instructunstatedsame——Memory (4-bit): 0.1 GB → 4.2 GB