Zero-Cost Interactive Tool

Estimate your website knowledge base & embedding tokens

See exactly how many vector chunks, 768-dimensional tokens, and crawl seconds your website requires before signing up. 100% computed in your browser.

100% Client-Side Simulation

Website Sizing & Token Estimator

Estimate your total vector embeddings, chunk volume, and recommended plan tier with zero network requests.

250 pages
10 pages1,500 pages3,000+ pages
600 words/page
100 (Short store item)600 (Standard)2,000 (Deep article)
20 document pages
0 pages150 pages300+ pages
TOTAL CONTENT

159,000

Extracted words across web & files

EMBEDDING TOKENS

211,470

768-dimensional vector representation

VECTOR CHUNKS

661

350-token searchable passages

EST. CRAWL TIME

~14 seconds

Sub-50ms Cheerio high-speed crawler

Recommended Plan
$59 / month

SiteMind Growth Plan

5 connected domains & 5,000 monthly credits

  • Sufficient capacity for your 661 chunks
  • 1:1 honest credit standard (zero dynamic model multipliers)

No credit card required. Full access to all Growth features during trial.

Embedding Math

How the token estimation engine works

01. 1.33 Token Ratio

In standard English text, 1 word represents approximately 1.33 tokens when processed through Google’s `gemini-embedding-001` model.

02. 350-Token Overlap Chunks

Documents are split into 350-token semantic chunks with 50-token contextual overlaps to preserve business rules and sentence boundaries.

03. 768-Dim Vector Coordinates

Each chunk is indexed into PostgreSQL `pgvector` as a 768-dimensional float32 coordinate, enabling sub-15ms semantic cosine similarity search.

Experience Real Sizing

Ready to test on your live website?

Paste your real URL, let SiteMind crawl your pages automatically, and see your actual live vector knowledge base in under 2 minutes.