$0.022 per 1M tokens.
There is no free API; it costs $0.022 per 1M tokens. The weights are open, so you can run it yourself.
#41 of 200 in Embeddings. Strongest at: Embeddings & Retrieval.
On MTEB German v1 · Classification · German, E5 Mistral 7B Instruct scores 72.5 vs 72.8 — about the same.
Released 2024-02-08. 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.
E5 Mistral 7B Instruct is an open-weights embeddings & retrieval model from intfloat, released Feb 2024.
Jina Embeddings v5 Text Nano (cheaper), F2LLM v2 8B (open weights).
CheaperJina Embeddings v5 Text NanounstatedOpen+18%-11%$0.02Max input: 33K tokens → 8K tokens+1cOpenF2LLM v2 8Bunstated+22%——Memory (4-bit): 4.3 GB → 4.5 GBiCheaperMultilingual E5 Large InstructunstatedOpen-4%-55%$0.01Max input: 33K tokens → 1K tokens
CheaperBge M3unstatedOpen-5%-55%$0.01Max input: 33K tokens → 8K tokens+1iCheaperMultilingual E5 LargeunstatedOpen-7%-55%$0.01Max input: 33K tokens → 1K tokens