Elevarq Analyzer pricing

Pricing for Elevarq Analyzer, the PostgreSQL intelligence platform. Starting at $249/month. Four packages. Annual billing by default (2 months free), monthly available. Self-hosted, airgap-friendly, license activation in minutes. Contact us to get started.

Flat pricing — no token billing, no usage caps

Plans

Pick the tier that fits your scale. Annual billing is the default — 2 months free; monthly is available too.

Annual billing is recommended (about 2 months free). Prices shown in USD for your region — display only; your final charge is set by the billing country you choose at checkout.

Starter

$208/ month

Billed annually — $2,490/year

Your private CPU AI DBA for a single PostgreSQL database.

1 database
PostgreSQL analysis
English-language analysis
CPU runtime
Community support

Professional

$416/ month

Billed annually — $4,990/year

GPU-powered fast analysis for teams running a small fleet.

3 databases included
PostgreSQL + TimescaleDB analysis
Analysis in 6 languages
GPU runtime
OIDC SSO
1 ticket integration (GitHub, GitLab, Jira, Linear)

Why Elevarq Analyzer

Elevarq is the paid PostgreSQL analysis product built to be right about your database. Its findings come from models purpose-trained for PostgreSQL, run locally on your own infrastructure, and matched to the exact version you run, so the advice fits your release instead of mixing in settings and features from versions you do not have. Every recommendation is anchored to an explicit rule set, checked against cited evidence before you see it, and handed to your team as an implementation-ready recommendation that you apply through your own change process. Elevarq cannot modify your database and never runs anything against it. A general-purpose assistant pointed at your database cannot make those guarantees: it has no version discipline, no grounded rule set, and no verifiable evidence trail. On a production database, accuracy is the product, and that is what you pay for.

Add-ons

Two levers, both adjustable at any time within the year.

Parallel Analysis

Do you run a fleet? This flat unlock (Professional and Business) lets one GPU analyze several databases at once instead of one report at a time — so every database stays continuously current, not just the one you had time for. Add it when you operate multiple databases and want them all kept up to date. Bounded by your GPU's memory; never metered per database.

+$299 / month

Flat per deployment.

Additional databases

Add databases beyond your plan's included count — Starter up to 3, Professional up to 10, Business up to 50. Enterprise database scale is contract-defined. Set up with our team; no tier-upgrade required.

+$50 / month per database

Flat per database. Discount when billed annually.

How pricing is scoped

Plans differ across four dimensions. Within each plan, the license file embeds the entitlements your Analyzer validates at runtime — there are no separate enforcement layers to manage.

Database count: Included: 1 / 3 / 25 (Starter / Professional / Business); Enterprise is contract-defined. With the +$50/database add-on you can scale to a maximum of 3 / 10 / 50; Enterprise is contract-defined. The Analyzer enforces your license's database limit — distinct databases past it are rejected.
Model tier: How deep the explanations go. Higher tiers (Small → XL, one per plan) give clearer, more thorough reasoning on the harder, subtler findings; most teams are well served by their plan's tier and step up only if they routinely hit complex, ambiguous cases. The model runs locally — your own dedicated instance, in your own environment.
Compute: How fast analysis runs. Starter uses CPU — thorough, but a full analysis takes a while, so it runs periodically. Professional and above use GPU — fast enough to run frequently and keep findings current. Choose GPU (Professional+) if you analyze often or run more than one database; the Parallel Analysis add-on then lets one GPU cover several databases at once.
Support level: Community / Business / Enterprise tiers. Differs on response SLA, escalation path, and whether air-gapped update packages are included.

No token-based billing. No usage caps. No surprise invoices.

Most AI-enabled tooling charges per token of LLM consumption. Elevarq Analyzer doesn't. Every customer runs their own local LLM in their own environment, against their own snapshots — there is no shared SaaS inference pipeline, no per-call metering, no cost we pass through. You pay for capability (databases, model tier, compute, support), not for how much you happen to use it.

The practical consequence: pricing is predictable, billing is monthly or annual at your choice, and a busy month doesn't cost more than a quiet one.

Billing and license activation

Monthly or annual, your choice. Annual billing is the same plan at a ~17% discount — approximately two months free per year. Switch between the two views with the toggle above the plan cards.
Signed license file. On purchase you receive an Ed25519-signed JSON license file. Customer ID, expiration, max databases, model tier, compute (CPU/GPU) flag, enabled features, and support level are all embedded; the Analyzer validates them on every run.
Activate offline. Upload the license file to your Workbench's setup wizard once. No internet required at activation time, no phone-home thereafter. Air-gapped deployments are a first-class case.
Anti-reuse enforced. The same license cannot be re-activated against a different Elevarq Analyzer deployment without re-issuance. The anti-reuse ledger lives in your Workbench's local store, so volume replication is detectable rather than silently licensed.

Not ready to buy? Start with Elevarq Signals.

Elevarq Signals — the open-source collector component — is free under BSD-3-Clause. Run it standalone against any PostgreSQL deployment with no Elevarq Analyzer license and no commercial commitment. The snapshot it produces is exactly what the commercial Analyzer consumes; you confirm the safety model and data-boundary hands-on before you commit to anything.

Evaluate Elevarq Analyzer with your own data

Before you buy, run the evaluation: deploy the open-source collector against a real database, analyze a real snapshot, and judge the tickets on your own workload. Plan about an hour.

Safe to deploy

The collector is open source, read-only by design, and needs only a least-privilege monitoring role — no superuser, no schema changes, no agents on the database host. It refuses to run with unsafe roles.

Your data stays put

Snapshots contain statistics, not query results, and stay in your network. You decide where the snapshot file goes; nothing is sent to Elevarq. The whole evaluation runs in your infrastructure.

About an hour

Collector up in minutes via Docker Compose or a binary install, first snapshot shortly after, then analysis with a 14-day evaluation license we issue by hand. One afternoon covers it comfortably.

What you get out

Findings against your own database: diagnosis, cited evidence, recommended actions, and remediation SQL — tickets you can hand to the team that ships the fix. You judge those, not a canned demo.

The proof pack is a reviewable bundle of representative findings — useful when you want to see real output before deploying anything. The step-by-step version of the evaluation lives in the docs: prerequisites and the one-hour flow.

Beyond Enterprise · genuine outliers

The four-plan structure covers the large majority of cases — we mean it when we say it. If your situation genuinely sits outside it (non-standard procurement clauses, fleet sizes well beyond the Enterprise tier, regulated negotiations your legal team needs to drive), reach out and we'll have an exception conversation. It is the exception, not the path most customers take.

Start an exception conversation →