Every objection. Every clarification. One searchable list.

No. Assess and Operate both offer monthly (cancel-anytime) and annual billing. Annual saves the real dollar gap per tier — $1,188/yr on Assess, $3,588/yr on Operate, about 20% off the monthly run-rate. Program is annual-only.

Most companies have one production environment. Charging more for "more environments" rewards the wrong thing. We tier on what you need — depth of audit, breadth of tool coverage, a re-check on every upload vs a one-time decision — not on how much you use us. Every paid tier includes unlimited environments.

Operate gives you 3 Configuration Reviews a year, each a dated, scoped read of any one tool you choose (firewall, identity provider, cloud IAM, and so on) — right if you have a handful of critical systems to scrutinize one at a time. Program covers all 37 supported tools with unlimited reviews, multi-vendor attack-path stitching, and proof-over-time. If you're running a full security program with identity, network, endpoint, cloud, email, and SIEM all needing ongoing scrutiny, Program is the fit. (Today you upload exports/configs and we re-parse them — live read-only system connections are on the post-launch roadmap, not live yet.)

Assess and Operate: yes — monthly billing cancels at the next billing cycle. Annual billing runs to term. Program is annual-only; cancellation takes effect at end of the 12-month term.

No. We don't offer refunds, but you can cancel before any future charge. After cancellation, your account is read-only for 90 days so you can export your data; we retain the data for another 90 days for re-activation; deleted permanently after 180 days.

You keep your data. You lose access to features the lower tier doesn't include. Your existing PDFs remain downloadable for 90 days. You can re-upgrade anytime.

Credit card (Visa, Mastercard, Amex). ACH for annual purchases over $10,000 — email billing@cybertwin.io to set up. Wire transfer for multi-year contracts. No invoicing on monthly plans.

No free trial. Card is required at signup. The sample report at /sample-report is the evaluation path — download it, see exactly what the engine produces, decide.

Monthly billing is charged every 30 days. Annual is one upfront payment and saves the real dollar gap per tier — $1,188/yr on Assess, $3,588/yr on Operate, about 20% off the monthly run-rate. Program is annual-only. Both options support the same feature set on the chosen tier.

Upgrade anytime — proration is applied automatically, you pay the difference for the remainder of the term. Downgrade takes effect at the end of your current billing cycle (you keep access at the higher tier through the term you paid for). No upgrade penalties, no downgrade penalties.

Yes. 10% off year 2 of an annual subscription. 20% off year 3. Three-year prepay available — email billing@cybertwin.io. The discount applies to the published list price; we don't negotiate beyond what's published.

30% off any tier for: registered 501(c)(3) nonprofits, accredited education institutions, pre-seed and seed-stage startups (under $5M raised). Email hello@cybertwin.io with proof before signup. Discount applies for as long as you remain eligible.

Renewal is at the then-current published list price. If our list price hasn't changed, your price doesn't change. On Program and Enterprise, each environment has a year-in-review report you can download before you renew: the configuration reviews run, the frameworks covered, and how many findings a later upload showed closed across your organization. Its dollar figures are estimates, shown as ranges and labelled modeled.

When you need any of: white-label PDFs (custom logo + cover branding), single-tenant deployment, multi-entity grouping (parent + subsidiaries under one contract), a dedicated success manager, SLA-backed response time with contractual penalties, custom framework mapping, or custom integrations. Email enterprise@cybertwin.io or use the contact form — pricing typically lands in the $72,000–$150,000/year range depending on scope.

Three steps, shown as a checklist on your home screen. Describe your environment (the intake wizard), add a real artifact (a config export; a diagram or a prior audit is read for follow-up questions), and open your risk readout. Each step unlocks the next; the first assessment lands in the same session.

Whatever you have. A tool configuration export (firewall, identity provider, cloud IAM) also runs a Configuration Review on every plan — 1 review a year on Assess, 3 a year on Operate, unlimited on Program and Enterprise — with no live scan, no agent, no connected integration. A network diagram (draw.io), a prior audit report or a vendor invoice is read to ask you follow-up questions; your answers, not the file, update the model. No artifacts yet? The intake wizard alone produces a scored assessment.

Assess and Operate include 5 team members. Program includes 10. Enterprise is unlimited. Seats cover the people working in the app — the PDFs and report links you produce can go to anyone.

Most teams finish the intake in a single sitting and see first findings within minutes. Your full report (recommended stack, ALE, compliance scoring, attack-path simulation, deployment plan, audit-evidence pack, board pack) is ready as soon as the engine finishes.

Yes — every paid tier is unlimited. You typically run one environment per business unit, region, or production-vs-staging boundary. Reports can be scoped to a single environment or compared across environments via the Environment Diff report.

Not yet — live read-only API connections (Microsoft Graph / AWS / FortiManager / Cloudflare etc.) are on the post-launch roadmap. Today you provide artifacts — config exports, identity/CSPM exports — and CyberTwin re-parses the real bytes; nothing is scanned live and no agent is deployed. When connections ship, they'll be read-only with scoped credentials.

All paid tiers include all 24 supported frameworks: SOC 2, ISO 27001, PCI DSS v4, HIPAA, GDPR, NIST 800-53, NIST CSF, CIS v8, NIS2, DORA, Cyber Essentials, CMMC L2, NCA ECC v2, NCA CCC, NCA CSCC, NCA OTCC, SAMA CSF, PDPL-KSA, PDPL-UAE, UAE IA / NESA, DESC, Australia Essential Eight, India DPDP, and Brazil LGPD.

The engine computes capability requirements deterministically from your industry, headcount, region, and frameworks; matches them against 99 catalog products with real prices; runs cross-vendor attack-path simulation across 14 vendor plugins; scores compliance against 24 frameworks simultaneously with explicit/derived/ceiling provenance. The output is auditable end-to-end — every recommendation cites its catalog source.

Yes, on every plan. You get three priced stacks from the intake, or a target architecture built from your requirements (AWS, Azure, on-premises or hybrid) and re-run through the attack-path engine. For the Balanced stack, the deployment plan puts every product in order, with a written setup guide (the settings, prerequisites and checks, filled in with your details) where one exists and vendor docs where it does not; a design built from requirements also comes with a draft FortiGate config and identity and cloud specs. Every guide and config is a draft: your team reviews, tests and applies it, and CyberTwin never changes anything on your systems. From Operate, each setup step is tagged with the compliance controls it satisfies. Designs and prices are labeled modeled.

Customer data at rest lives in the EU — Frankfurt, Germany. Data is encrypted in transit (TLS 1.2 or higher) and the underlying volume is encrypted at rest. Tenant isolation is at the database row level. Audit logs are retained for 12 months (searchable in-app and exportable as CSV over that window). The full subprocessor roster and locations, including the US-based ones, are on /trust.

Permanent hard-delete is by request on every plan: from Settings → Data management (or email trust@cybertwin.io) send a "Deletion request" from an Owner/Admin address — we execute it within 5 business days and confirm by email. You can also export your full tenant data yourself, any time, from the same page.

No. We do not train models on your data, and under Anthropic’s Commercial Terms Anthropic does not train on content sent through its API. The findings, attack paths and scores come from a deterministic engine, not an AI model. Some features do call Anthropic’s Claude API (US). What is sent: profile details (industry, size, region, frameworks, budget ceiling) and the engine’s structured output, to write the report narrative; the free-text answers and questions you type; the text and tables of documents you upload for document analysis, and the draw.io XML of a diagram you ask us to read, with secrets redacted first; and, only if you switch on AI commentary for a configuration review (it is off by default), a short digest of its findings (check id, severity, title, config section), never the configuration file itself. Anthropic’s API keeps inputs for up to 30 days under its default policy; we have no zero-retention agreement.

Known secret patterns across 8 secret classes (API keys, connection strings, passwords, tokens, certificates, PSKs, private keys, generic secrets) are redacted in memory before anything is stored. For a configuration review, the redacted text is what gets saved. For an uploaded document, secrets are removed from the extracted text before it is saved or sent to the AI model, and the original file is never stored. The one exception: you can switch redaction off for a configuration review by ticking an explicit acknowledgment, and then the text is saved as uploaded.