Tools•calculator
OpenAI Embeddings for Embeddings & Vector Search
Calculate estimated storage requirements for embedding vectors based on dataset size, dimension, and data type.
Try the tool
client runnerStorage Estimate
Run the tool to see output.
Examples
100k vectors at 768 dimensions (float32)
{
"vector_count": 100000,
"dimension": 768,
"data_type": "float32"
}Expected output
307.2 MB
50k vectors at 2048 dimensions (float16, compressed)
{
"vector_count": 50000,
"dimension": 2048,
"data_type": "float16",
"compression": true
}Expected output
102.4 MB (compressed)
How it works
The calculator estimates storage by multiplying the number of vectors by the embedding dimension and the bytes per value (4 for float32, 2 for float16/bfloat16). Compression reduces storage by 50% if enabled.
.