Pricing that makes abuse prevention a no-brainer.
Every Dregs plan includes the core tools for catching fake signups, freeloaders, and abuse: device fingerprinting, identity scoring, behavioral analysis, notifications, webhooks, and more. Pick the plan that matches your team's scale and how much control you want over the scoring.
Starter
For early-stage teams getting serious about fraud and abuse.
Includes 100 active identities
+ $0.06 per additional active identity
- 500,000 checks/month
- 3 team members
- 1M events/month
- 10 badge rules
- 10 escalation rules
- 1 channel
- 90-day event history
Growth
For scaling teams with more traffic and a need for deeper analysis.
Includes 5,000 active identities
+ $0.05 per additional active identity
- 2,500,000 checks/month
- 5 team members
- 5M events/month
- 20 badge rules
- 20 escalation rules
- 3 channels
- 180-day event history
Advanced
For sophisticated teams wanting extra control over identity scoring.
Includes 10,000 active identities
+ $0.04 per additional active identity
- 5,000,000 checks/month
- 10 team members
- 10M events/month
- 30 badge rules
- 30 escalation rules
- 5 channels
- 365-day event history
- Custom analyzers
- Custom datasets
All plans include a 14-day free trial. No credit card required.
Need higher limits or a custom plan? Get in touch. Want help with setup and tuning? See Services.
Continuous scoring, priced by identity
Many fraud products charge per evaluation: you call them at signup, at login, or before an access decision, and each call returns one verdict about one moment. If you want to know whether the same account looks worse a week later, that's another evaluation.
Dregs charges per active identity instead. An account that was tracked and scored in the billing month counts once, and Dregs keeps scoring it across Humanity, Authenticity, Uniqueness, and Behavior as its events arrive. When the picture changes, badges, escalations, and webhooks follow, and you can train the scoring on accounts you already trust. Your bill tracks the number of accounts Dregs is watching for you that month, however often their scores change. See how Dregs works for the scoring model, Dregs vs Rupt for a product that meters evaluations, or Dregs vs ShieldLabs for one that bills per identification.
Each plan includes a monthly allowance of checks. A check is a score lookup for an identity you already track, through the API, the dashboard, or MCP. Dregs keeps scoring that identity as its events arrive, so the scores are already computed and a check reads them. Some products call that lookup an evaluation and charge for each one. On Dregs those lookups are included, and another score lookup is not another charge. Authenticity Check API calls are separate: a fresh scoring pass, including on an email you haven't tracked, counts against the monthly operations quota.
Every plan includes
Frequently asked questions
How Dregs scores each account describes how the scores are computed. The investigative and rule-tuning parts are usable entirely through your AI assistant or a coding agent such as Claude Code, Codex, or Cursor.
What is an active identity?
An active identity is one that was tracked and scored in the month. If a user of your application signs up in January but doesn't return in February, they don't count toward February's limit. Inactive identities are still kept in Dregs, but you only pay for the ones that were tracked and scored during the month.
What happens if I exceed my active identity limit?
If you exceed the included identities in your plan, you're billed for the overage at your plan's per-identity rate. This is calculated automatically at the end of each billing cycle. You can monitor your usage in the dashboard at any time.
How is this different from products that charge per evaluation?
Per-evaluation products charge for a discrete check at a moment you choose, such as signup, login, or an access decision, and checking the same account again later is another evaluation and another charge. Dregs charges per active identity instead. An account that was tracked and scored in the billing month counts once, and Dregs keeps scoring it across Humanity, Authenticity, Uniqueness, and Behavior as further events arrive. A later score lookup for an identity you already track is a check, and each plan includes a monthly checks allowance for those lookups. Another score lookup is not another charge. Authenticity Check API calls are separate: a fresh scoring pass, including on an email you have not tracked, counts against the monthly operations quota. Badges, escalations, and webhooks follow when the scores move, and you can train the scoring on accounts you already trust. What you pay for is the number of accounts Dregs is watching for abuse that month.
What is a check?
A check is a score lookup for an identity you already track, through the API (GET /api/identities/{id}/scores), the dashboard, or MCP. It reads the current Humanity, Authenticity, Uniqueness, and Behavior scores, which Dregs has already computed as that identity's events arrived. Some products call the same lookup an evaluation. Each plan includes a monthly checks allowance. The subscription is still priced by active identity, so another score lookup for an account you're already watching is not another charge. Authenticity Check API calls are separate. A call such as POST /api/tools/authenticity-check runs a fresh scoring pass, including on an email you have not tracked, and counts against the monthly operations quota.
What counts as an event?
An event is any tracking call sent to Dregs from your integration. This can include pageviews, signups, logins, form submissions, or other user actions. Each identification or tracking call counts as one event.
What are the plan limits?
Event allowances, included identities, API rate limits, payload caps, and webhook retries are listed on the limits and quotas page in the manual.
What are custom analyzers and datasets?
Custom analyzers let you write your own JavaScript scoring functions that run alongside Dregs's built-in analyzers. They have access to the identity's events, devices, and datasets, and can implement domain-specific fraud detection logic unique to your application. Custom datasets are lists or mappings stored in Dregs and available to your analyzer functions, such as a list of banned domains or bad IPs that your own rules can check against.
Can I change plans later?
Absolutely. You can upgrade or downgrade your plan at any time through your account settings.
Do you offer annual billing?
Sure, we can offer annual billing for custom plans. Please contact dregs@dregs.com if you'd like to discuss annual pricing or other custom plans.
What happens when my free trial ends?
If your free trial ends and you haven't entered a payment card or chosen a paid plan, Dregs will pause scoring and analysis for your account. Your data is preserved for a little while, and during this grace period events continue to be ingested so you don't lose any tracking data. You can upgrade to any plan to resume identity scoring instantly. If you don't upgrade to a plan, your account will be shut down after the grace period.
Why not build this ourselves?
You could probably stand up an initial version in a few days or weeks, but fine-tuning the rules and finding all the edge cases and abuse vectors will take months or years (as the Dregs team knows first-hand). Dregs keeps the disposable email domain list current, keeps device and network signals current as events arrive, and links related accounts across signups. When behavior changes, the scores move with it, so an account that looked fine at signup can still look suspicious later. For most teams, letting Dregs handle it frees you to focus on your actual product.
Can't we just ask an AI to score each signup?
You can, for a handful of accounts. At signup volume, each model call adds cost and seconds, the verdict can change between runs, and the model only sees one record, not the device, IP, and email links across your users over time. Dregs keeps a consistent score for each account as events arrive, and you can still bring an assistant or a coding agent such as Claude Code, Codex, or Cursor in through the MCP server to explain it.