How we score and check trading bots
One formula for every platform, one scale out of 10, and a primary source behind every fact.
The score
score = Σ (criterion score × weight) ÷ Σ (weights of criteria that have a source)| Criterion | Weight | What we measure | Accepted sources |
|---|---|---|---|
| Bot variety | 25% | Strategies offered, configurability, presets | Platform documentation, hands-on test |
| Fees | 20% | Bot subscription and trading fees vs peers | Official fee schedule |
| Safety & licensing | 20% | Licences, registrations, proof of reserves, incidents | Regulator registers, platform disclosures |
| Ease of use | 15% | Setup steps, mobile app, defaults | Hands-on test with date |
| Support | 10% | Channels, languages, hours | Platform support pages |
| Transparency | 10% | Published backtest and fee methodology | Platform documentation |
How facts become points
Two criteria are calculated from sourced facts with a fixed formula, so anyone can repeat the result:
- Bot variety: 0.75 points per strategy family (up to 8), +2 if futures bots are offered, +1 for ready-made or AI-suggested presets, +1 for an official mobile app. Maximum 10.
- Fees, exchanges with free bots: 10 points at an average spot fee of 0.05%, minus 0.4 for every 0.01 percentage point above it (0.10% gives 8).
- Fees, crypto bot platforms: 8 minus the cheapest monthly price ÷ 10, plus 1 if a free plan exists. Exchange trading fees come on top, so these platforms start lower.
- Fees, Telegram and on-chain bots: 10 minus 5 × the highest per-trade fee in percent (1% per trade gives 5).
- Fees, stock, options and forex tools: free gives 10; a free tier plus paid plans gives 9; a yearly advisory fee gives 10 minus 10 × the percentage (0.25% gives 7.5); a subscription gives 10 minus the cheapest monthly price ÷ 15.
Fee formulas differ by section because costs work differently, so compare scores within a section, not across sections.
Safety, ease of use, support and transparency are scored only after a regulator-register check, a dated hands-on test and a support check. Until then they show “Not scored” and every score is marked preliminary.
Rules we don’t break
- 01No source, no scoreA criterion without a primary source is not scored; its weight is redistributed and the score is marked preliminary.
- 02Unknown is not zeroMissing values show “Not verified”, never 0, “no” or “available”.
- 03One scaleEvery score is out of 10. We never show stars next to it.
- 04Two different dates“Score last changed” and “Review updated” are shown separately.
- 05Ads don’t rankPaid placements are marked AD and keep the position their score gives them.
- 06Lists by ruleRankings are built from a published rule, ordered by score, and need at least three platforms.
How we check availability by country
- Only primary sources: the platform’s terms, restricted-country lists, help center, and regulator registers.
- Every rule gets a link, a short quote and the date checked.
- If the source is silent about a country, the status is “Not verified”, never “Available”.
- A new check replaces that country’s rules completely and leaves other countries untouched.
Who writes and how we make money
Reviews are written and checked by the BotIndex editorial team. We earn affiliate commissions on some “Visit” links and sell clearly marked AD placements. Neither changes a score or a position in a ranking.