🎯 MTEB #1 Vector Embeddings

Voyage AI Pricing & Rates

Domain-specialized state-of-the-art embedding and reranker models. Top-tier retrieval accuracy with 200M free tokens on sign up.

voyage-3-lite / 1M

$0.020

512 dims, ultra fast

voyage-3 (General) / 1M

$0.120

1024 dims, SOTA MTEB

voyage-code-3 / 1M

$0.120

Code search specialist

Free Registration Tier

200M

Tokens free credit

🧮 RAG Knowledge Base Embedding Cost Calculator

Calculate total embedding ingestion cost and Pinecone/Qdrant vector RAM storage overhead:

Documents / Articles to Ingest 25,000 documents

Average Document Length 1,500 tokens / doc

Total Corpus Tokens

37.5M

voyage-3 Ingestion Bill

$4.50

voyage-3-lite Bill

$0.75

Within Free Tier?

100% Free

📋 Voyage AI Complete Model Catalog

Model Name	Dimensions	Context Window	Price / 1M Tokens	Primary Use Case
voyage-3	1024	32,000	$0.120	General domain high-precision RAG
voyage-3-lite	512	32,000	$0.020	Latency-critical apps & massive corpora
voyage-code-3	1024	32,000	$0.120	Repo search, code QA, symbol retrieval
voyage-law-2	1024	16,000	$0.120	Case law, statutory contracts & discovery
voyage-finance-2	1024	32,000	$0.120	10-K filings, earnings calls, market data
rerank-2	-	16,000	$0.050 / 1k queries	Two-stage cross-encoder re-ranking

💻 Quickstart with Voyage AI Python Client

import voyageai vo = voyageai.Client(api_key="your_voyage_api_key") # Generate dense vector embeddings for documents docs = ["CheckAPICost tracks real-time AI API pricing.", "Voyage-3 achieves MTEB #1 ranking."] result = vo.embed(docs, model="voyage-3", input_type="document") print(f"Vectors generated: {len(result.embeddings)}, Dimensions: {len(result.embeddings[0])}")

Frequently Asked Questions

How much does Voyage AI cost for vector embeddings?

Voyage AI offers voyage-3-lite at $0.020 per 1M tokens, while flagship domain models (voyage-3, voyage-code-3, voyage-law-2, voyage-finance-2) are priced at $0.120 per 1M tokens.

Does Voyage AI have a free tier?

Yes, Voyage AI grants 200 million free tokens upon registration, which is sufficient to index up to 150,000 PDF documents or thousands of code repositories without paying.

How does Voyage AI compare to OpenAI text-embedding-3-large?

Voyage AI consistently outranks OpenAI text-embedding-3-large across Hugging Face MTEB benchmarks while being 7% cheaper ($0.120 vs $0.130/1M tokens) and requiring smaller vector dimensions (1024 vs 3072 dims), cutting vector database RAM storage by 66%.