Sova IntelSova Intel
Back to Help
Feature Guide

Similarity Lab

Compare wallets by overlap and capital allocation to uncover clusters, mirrors, or coordinated behavior.

Why this exists

This turns a wallet list into relationship intelligence instead of isolated profiles.

Best for

  • Testing whether a wallet cluster really trades alike.
  • Checking overlap between top holders or tracked cohorts.
  • Spotting copy-trading, shared themes, or portfolio convergence.

Open the feature

Use the guide for framing. Use the live page for actual analysis.

Open Similarity Lab

Screenshot

Main product view for Similarity Lab.

Similarity Lab dashboard screenshot

Quick read

Read workflow first to understand how the page is supposed to be used.

Read what to expect and limits before you trust the output literally.

Use the FAQ for edge cases.

Read next

Tracked WalletsHolder Profiles

Typical workflow

1

Paste at least two wallets, or start from a token and select holders.

2

Queue the async job and wait for the comparison result.

3

Use the strongest pairs, clusters, and common tokens to decide what matters next.

What data is inside

Pairwise similarity scores and shared-token overlap.

Wallet-level balance context and common-token views.

Cluster-style grouping signals and global metrics for the full set.

What to expect

This is an async job and larger wallet sets take longer.

Known system or non-trader wallets lower the usefulness of the comparison and can be skipped.

Metadata and pricing enrichment can continue after the main similarity result loads.

Limits

Minimum 2 wallets, maximum 30 wallets per run.

Credit cost is dynamic: 10 if wallets are already synced, 20 if sync is needed.

Heavy result payloads are job-backed rather than fully inline.