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.
#190 of 200 in Embeddings.
On MTEB Portuguese v1 · Reranking · Portuguese, Distilrubert Small Cased Conversational scores 59.8 vs 62.2 — lower.
Released 2022-06-28. It is the current version in its line.
Yes, the weights are open (not specified licence).
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Distilrubert Small Cased Conversational is an open-weights embeddings & retrieval model from DeepPavlov, released Jun 2022.
EmbeddingGemma 300M (better), E5 Mistral 7B Instruct (open weights).
BetterEmbeddingGemma 300MunstatedOpen+43%—$0.002Languages: not verified+1iOpenE5 Mistral 7B Instructunstated+46%—$0.022Memory (4-bit): 0.1 GB → 4.3 GBiBetterMultilingual E5 Large InstructunstatedOpen+44%—$0.01Memory (4-bit): 0.1 GB → 0.3 GB
BetterQwen 3 Embedding 0.6BunstatedOpen+43%—$0.01Languages: not verified+1iBetterMultilingual E5 LargeunstatedOpen+41%—$0.01Memory (4-bit): 0.1 GB → 0.3 GB
Snowflake Arctic Embed L v2.0unstatedOpen+44%——Memory (4-bit): 0.1 GB → 0.3 GBiMultilingual E5 Large InstructunstatedOpen+44%—$0.01Memory (4-bit): 0.1 GB → 0.3 GB