Methodology - How We Research and Rank

Data first, prose second

Every provider entry is a structured fact set: vertical and type, price and unit, minimum, KYC, payment methods, contract term and payout terms, proof of hashrate/compute (pool-worker visibility, on-chain payouts, attestation), operator entity and jurisdiction, and dated sources. Reviews are AI-drafted from that fact set only - the model may not invent numbers, terms or history. Drafts are fact-checked against the database before publication, and every fact row carries a verification date.

Units are normalised per vertical

Mining contracts are priced per TH/s (or GH/s) per day plus a maintenance fee; hosting per $/kWh; GPU compute per GPU-hour; storage per TB-month or per GB once; naming per year; staking per commission percentage. The comparison table shows each provider's own unit, never a fake common one.

Scoring

Editorial scores (0-10) use seven dimensions - economics, reliability, transparency, track record, contract terms, ease of use and support - weighted differently per vertical. For mining, for example: economics 22%, transparency 20%, reliability 18%, track record 15%, terms 13%, ease 7%, support 5%. A storage provider is judged on different weights (ease and reliability weigh more). User ratings are shown separately, smoothed with a statistical prior, and never mixed into editorial scores.

Proof and red flags

"Proof: published" means we found a verifiable signal - a public pool worker, on-chain payout addresses or hardware attestation - not a marketing claim. Red flags (guaranteed returns, referral pyramids, unresolved payout complaints, criminal cases) are logged with sources and drive the Scam Watch status.

Freshness

Prices and availability change fast. We run daily reachability checks on every listed domain, log public network/hashprice and GPU-price snapshots, and regenerate reviews on a cycle plus on demand when facts change. Every review shows its version, generation date and fact-verification date.

Independence

Commercial relationships never influence scores, ranks or the scam list. Providers can submit factual corrections with sources via contact; we verify and correct within 7 days. We do not remove dead or scam listings in exchange for anything.