Nomic Embed v2
Free.ai (self-hosted)
·
embeddings
·
~100 Koffice FilterName call
Nomic Embed v2 is an embeming model gebou deur Nomic AI. Sterkste by Retrieval augmented generation with flexible vector sizes.. Self-hosted on Free.ai GPUs — runs free against your daily token pool (100 tokens per roep). Released under Apache 2.0 — commercial use permitted on Free.ai.
Gebruik via API
OpenAI- versoenbaar met REST API. Genereer 'n sleutel en noem hierdie model in sekondes.
curl -X POST https://api.free.ai/v1/image/generate/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"model":"nomic-embed-v2","prompt":"your prompt here"}'Vrae wat dikwels gevra word
Nomic Embed v2 converts text into a dense vector (a list of floats) that captures meaning. Use it for semantic search, clustering, recommendation, retrieval-augmented generation (RAG), and any task where "is this text similar to that text" matters.
Tipiese dimensies is 384, 768, 1024 of 1536 afhangende van die model. BGE-M3 uitstuur 1024-dim; OpenAI Ada stel 1536 vry. Die API-reaksie sluit die dimensie in sodat jou vektor DB die regte indeks kies.
Modern embedding models (including most options on Free.ai) are trained on 100+ languages. Cross-language retrieval works — search in English, match documents in Spanish.
512 tot 8 192 inligting na gelang van die model. Lang dokumente word langer afgeknot in paragrawe voor opvou.
Nomic Embed v2 loop op ons eie GPU's en is een van die goedkoopste gereedskap wat omstreeks ~100 seine per oproep van jou daaglikse vry poel getrek is. $5 = 200K ides.
Ja 98 POTT 'n lys stringe na /v1/embedings/ en Nomic Embed v2 gee terug' n lys van vektore in dieselfde volgorde. Bochtch grootte na 2 048 per versoek.
L2-ormaaliseer deur verstek takiesl cos-ooreenkoms = punt produk. Pass takiesialise=SvalsName
Enige ą Pinecone, Weiviate, Qdrant, Chroma, pgvector, FAISS, LanceDB. Nomic Embed v2 keer terug gewone JSON dryf; die DB sien nooit die model nie.
Ja 98 POST na /v1/embedings/ met model="Nomic Embed v2". OpenAI- versoenbaarte reaksie, so bestaande kliënt biblioteke werk onveranderd. /api/ het die volledige verwysing.
Self-gehostde modelle hou jou teks op ons GPUs en gooi dit weg na die oproep terug. Premium gaan deur met 'n DPA. Ons oefen nie op jou invoere nie.
Sub-100ms for short text on self-hosted, 100–500ms on premium. Batch calls scale roughly linearly — 1,000 chunks complete in 2–10 seconds.
Yes — Free.ai grants commercial use of embeddings. Build production search, RAG pipelines, recommendation systems with no per-vector royalty.