Budgeting for dbt Cloud comes down to two levers: developer seats and usage. Seats are straightforward; usage is measured as successful models built in deployment runs and queried metrics in the Semantic Layer. This quick dbt Cloud pricing guide explains those meters so your estimates are accurate, then walks through current plan entitlements, the Starter billing formula, and two worked bills you can benchmark against.
You’ll also get a quick Starter vs Enterprise decision check, as well as practical cost-control tips.
Explore Our dbt Services
How dbt Cloud Pricing Works (In 30 Seconds)
dbt Cloud charges for seats and usage. Usage has two meters:
- Successful Models Built (scheduler, CI, or API-triggered runs in deployment environments), and
- Queried Metrics if you use the Semantic Layer. Development IDE runs don’t count; tests, seeds, ephemeral models, and snapshots also don’t count.
On dbt’s pricing page, the Developer plan is one seat and one project with up to 3,000 successful models/month (jobs pause beyond the limit), suitable for individual evaluation, not teams. the Starter plan is $100 per user/month and is designed for up to five developer seats, with 15,000 successful models/month, 5,000 queried metrics/month, and one project. Enterprise tiers raise those entitlements (see next section).
For Starter, dbt even gives a back-of-the-envelope estimator:
( $100 × # developer seats ) + ( (models built − 15,000) × $0.01 )
(charged monthly; usage in arrears).
dbt Cloud Pricing Plans and What You Actually Get
Developer (free)
- One developer seat, 3,000 models/month. If you exceed the cap, further runs are paused until next month or you upgrade.
Starter ($100 per user/month)
- Up to 5 developer seats, 15,000 models/month, 5,000 queried metrics/month, 1 project. API access and basic Catalog/Semantic Layer are included. Overages on models are billed monthly per the formula above.
Enterprise (custom pricing)
- Custom seat count, 100,000 models/month, 20,000 queried metrics/month, 30 projects; adds advanced Catalog/Semantic Layer, Mesh, Insights, and cost management.
Enterprise+ (custom pricing)
- Everything in Enterprise plus features for locked-down environments like PrivateLink and IP restrictions, plus rollback and hybrid projects; projects are unlimited.
Security/identity gating: SSO (SAML) and SCIM are Enterprise features; PrivateLink and IP restrictions are Enterprise+ (and certain Enterprise tiers). If you need those controls, you’re almost certainly shopping Enterprise.
For an in-depth, up-to-date feature comparison, see the dbt Cloud pricing page; this guide concentrates on what you’ll actually pay.
You May Also Like: dbt Semantic Layer at Scale: How to Build a Single Source of Truth for Enterprise Metrics
Usage That Drives Your Bill: What Counts (and What Doesn’t)
Because dbt Cloud bills on seats + usage, and usage = successful models built + queried metrics, here’s exactly what increments the meter, so the math below is accurate.
Counts (billed)
- Scheduler, CI (pull request), or API-triggered runs in a deployment environment
- Per-model success: if a 100-model job fails after 51 models, 51 count
- Semantic Layer: each successful request counts per metric in the request
Doesn’t count (not billed)
- IDE development runs
- Tests, seeds, snapshots, ephemeral models
- Semantic Layer metadata-only calls
We use this definition in the examples that follow and in the Starter estimate:
($100 × developer seats) + ((models built − 15,000) × $0.01)
Pro tips: keep PR/CI jobs lean, use state-based selection to build only changed models, and avoid unnecessary full-refreshes.
Two Worked Examples (So You Can Actually Budget)
Example A: Small Team on Starter
Team: 7 developers
Activity: ~50,000 successful models/month
- Seat cost: 7 × $100 = $700 / month
- Model overage: 50,000 − 15,000 = 35,000 models → 35,000 × $0.01 = $350 / month
- Estimated monthly total: $700 + $350 = $1,050
- Estimated annual: $1,050 × 12 = $12,600
This uses dbt’s own Starter formula and “what counts” rules. If your monthly models swing a lot, expect the bill to swing too because overage is metered monthly.
Gut-check: If you regularly exceed ~115k–130k models/month with a small team, Starter’s overages may start to feel spicy—talk to sales about Enterprise entitlements, or tune your jobs (see tips below).
Example B: Mid-Size Team on Enterprise
dbt’s UK G-Cloud price list includes illustrative enterprise math:
- 20 developers at $300/seat/month → $72,000/year
- Scheduler: 500,000 models/year at $0.01/model → $5,000
- Semantic Layer: 20,000 queries/month (240,000/year) at $0.075/metric → $18,000
- Illustrative annual total: $95,000
Treat this as ballpark list pricing; real enterprise contracts vary (discounts, commits, and included entitlements apply).
dbt Cloud Pricing Plans: Starter vs Enterprise – Which One Is Right?
Choose Starter if…
- You have ≤5 active developers and can keep builds ≲ 15k/month or are okay paying occasional overages.
- You don’t need SSO/SCIM, PrivateLink, or IP restrictions yet.
Choose Enterprise/Enterprise+ if…
- You need SSO/SCIM for governance, PrivateLink/IP restrictions, or regional controls.
- You expect sustained build volume near or above 100k models/month or heavy Semantic Layer usage.
- You want features like Mesh, Insights, and cost dashboards for platform-scale operations.
Not sure? B EYE can model your actual build pattern from scheduler and CI logs and tell you where the true break-even sits for your team.