Pricing

Simple. Honest.

One plan. Every platform. No per-comment fees, no token quotas, no "Enterprise — call us".

Professional
Everything you need from day one
£179.99
per month +VAT · or £1,799.90/yr
Annual = 10 months for the price of 12
1 brand entity included (Facebook + Instagram; TikTok & YouTube as we launch)
Extra entities +£143.99/mo each (80% of base plan)
5 authorised vetters included (+£5/mo each extra)
Per-tenant Telegram bot (your branding)
Calibration dashboard with agreement metrics
Audit log preserved forever
Hide-first invariant — public never sees unvetted
AI scoring with tenant-specific topic rules
Daily Telegram digest of activity
Override-as-signal alerts for tuning
Encryption at rest (AES-256-GCM)
GDPR/UK-GDPR data deletion on request
Start your 7-day free trial

Card up-front · No charge for 7 days · Cancel any time

Honest list

What we don't have, and why.

  • X / Twitter coverage

    X requires a paid API tier just to read or hide replies. The cost and access restrictions make it impossible to offer reliably. We've made the honest call to drop it rather than offer something we can't guarantee.

  • Threads, Bluesky, Mastodon coverage

    Those platforms don't expose hide or delete to third parties. We could file removal reports, but reports without action feel weaker than honest 'not covered yet'. Frank suggestion: if you're concerned enough about comments to be reading this page, we'd advise sitting out Threads, Bluesky and Mastodon for now. They lack the moderation controls your audience expects you to have.

  • Per-comment pricing

    A bad-actor swarm shouldn't spike your bill. One flat number means we're aligned — we want spam dealt with as much as you do.

  • A bigger Enterprise tier with sales calls

    If you need something special (white-label bot, dedicated infra, custom SLA), email us. No fake tiers.

Common questions

Five things people ask before signing up.

Does the AI ever post comments on my behalf?
No. Velvet Rope is read-and-moderate only. The AI scores incoming comments and our pipeline either hides them, deletes them, or surfaces them to your team in Telegram — it cannot write or post anything outbound. Your team always controls what stays public.
What happens when I cancel?
Your platform tokens are revoked, your data is deleted within 30 days. You can also revoke our app from your Facebook settings any time — that's the kill switch.
What if the AI gets it wrong?
Every auto-action shows up in Telegram with Right/Wrong buttons. If the AI auto-approved something it shouldn't have, tap Wrong → we delete the comment immediately. If the AI auto-deleted something it shouldn't have, tap Wrong → we record it as a training signal so the AI learns, but the comment itself can't be restored (Facebook's API has no undelete). Either way, calibration metrics track agreement rate and surface in your dashboard, so you can tune the thresholds when you spot a pattern.
Where is my data?
On infrastructure controlled by Geeky Bee Limited (UK), hosted in EU (Germany). Encrypted at rest. Per-tenant Postgres Row-Level Security so even our own application code can't read another tenant's data without explicit context-switching.
Can I add my own 'always human review' topics?
Yes. During onboarding we ask what subjects should never be auto-actioned by the AI — legal matters, deceased family, ex-partners, anything sensitive to your situation — and we bake those phrases into your account's AI prompt. The AI then forces every comment touching those topics to your team for human review, regardless of how positive or hostile the rest of the content reads.

Every comment is seen by the person who wrote it — that's how social platforms work, and it's not something we change. What we do is decide, quietly and quickly, whether it belongs in the room with your audience.

Genuinely warm comments — the kind that reflect the way you or your passion makes people feel — pass straight through without a hand being raised. Everything else is held at the rope until someone on your team has a look. Some are let in. Some aren't.