Which Workload Type Do I Need?
Lakemeter supports 14 workload types covering the full Databricks platform. This page helps you choose the right one for your use case.
A Lakemeter estimate showing multiple workload types with their individual costs and configuration summaries.
Animated: adding a new workload to an estimate — choose the workload type, configure compute and usage, then save.
Quick decision guide
| I need to... | Use this workload | Guide |
|---|---|---|
| Run scheduled ETL/data pipelines | Jobs | Jobs Compute |
| Work interactively in notebooks | All-Purpose Compute | All-Purpose Compute |
| Build declarative data pipelines with quality checks | DLT | Delta Live Tables |
| Run SQL queries, dashboards, BI tools | DBSQL | DBSQL Warehouses |
| Deploy a custom ML model as an API | Model Serving | Model Serving |
| Store and query vector embeddings (RAG) | Vector Search | Vector Search |
| Call open-source LLMs (Llama, DBRX, Gemma) | FMAPI Databricks | FMAPI — Databricks |
| Call commercial LLMs (Claude, GPT, Gemini) | FMAPI Proprietary | FMAPI — Proprietary |
| Use a managed PostgreSQL database | Lakebase | Lakebase |
| Host a web app on Databricks | Databricks Apps | — |
| Parse documents with AI | AI Parse (Document AI) | — |
| Generate images with AI | Shutterstock ImageAI | — |
Workload categories
Compute Workloads
These workloads run Spark clusters or SQL warehouses. Costs are based on instance types, cluster size, and hours of usage.
| Workload | Pricing model | Supports Classic | Supports Serverless | VM costs |
|---|---|---|---|---|
| Jobs | DBU/hr x hours | Yes | Yes | Classic only |
| All-Purpose | DBU/hr x hours | Yes | Yes | Classic only |
| DLT | DBU/hr x hours | Yes | Yes | Classic only |
| DBSQL | DBU/hr x hours | Yes (Classic/Pro) | Yes | Classic/Pro only |
AI/ML & Data Services
These workloads are always serverless with no VM costs. Pricing is based on GPU type, token volume, or provisioned capacity.
| Workload | Pricing model | Unit | VM costs |
|---|---|---|---|
| Model Serving | DBU/hr by GPU type | Hours | Never |
| Vector Search | DBU/hr by endpoint units | Hours + storage | Never |
| FMAPI Databricks | DBU per million tokens or DBU/hr (provisioned) | Tokens or hours | Never |
| FMAPI Proprietary | DBU per million tokens | Tokens | Never |
| Lakebase | DBU/hr (compute) + DSU (storage) | Hours + GB | Never |
| Databricks Apps | DBU/hr by app size | Hours | Never |
| AI Parse | DBU per 1000 pages | Pages | Never |
| Shutterstock ImageAI | Fixed per image | Images | Never |
Common fields
All workloads share these base fields:
| Field | Description |
|---|---|
| Workload Name | A descriptive label (e.g., "ETL Pipeline", "Analytics Warehouse") |
| Workload Type | The Databricks service type from the dropdown |
| Display Order | Sort position in the estimate (drag to reorder) |
Usage input modes
Compute workloads (Jobs, All-Purpose, DLT, DBSQL) support two ways to specify usage:
Run-Based (default): Specify runs per day, average runtime, and days per month. The formula is:
Hours/Month = (Runs Per Day x Avg Runtime Minutes / 60) x Days Per Month
Direct Hours: Enter total hours per month directly. Common values:
- 730 = 24/7 (always-on)
- 176 = 8 hours x 22 business days
- 44 = 2 hours x 22 business days
Model Serving, Vector Search, and Lakebase use direct hours only. FMAPI workloads use token volume (millions/month) or provisioned hours instead of either mode.
Pricing tier restrictions
Some workloads require a Premium pricing tier or above:
| Workload | Standard tier | Premium tier |
|---|---|---|
| Jobs (Classic) | Yes | Yes |
| Jobs (Serverless) | No | Yes |
| All-Purpose (Classic) | Yes | Yes |
| All-Purpose (Serverless) | No | Yes |
| DLT | Yes | Yes |
| DBSQL (Classic/Pro) | Yes | Yes |
| DBSQL (Serverless) | No | Yes |
| Model Serving | No | Yes |
| Vector Search | No | Yes |
| FMAPI Databricks | No | Yes |
| FMAPI Proprietary | No | Yes |
| Lakebase | No | Yes |
| Databricks Apps | No | Yes |
| AI Parse | No | Yes |
| Shutterstock ImageAI | No | Yes |
If you select Standard tier for your estimate, workloads that require Premium will not be available.
Cost components
Depending on the workload type, your total cost may include one or more of:
- DBU Cost -- Databricks Units consumed, multiplied by the $/DBU rate for your SKU
- VM Cost -- Cloud infrastructure costs for driver and worker VMs (Classic compute only)
- Storage Cost -- Provisioned storage (Lakebase DSU, Vector Search overflow)
- Token Cost -- For FMAPI workloads, based on millions of tokens processed
The total monthly cost for each workload is displayed in the estimate summary. All workloads sum to the estimate total.