Lightdash vs Holistics (2026): Pricing, Semantic Layer and Embedding Compared
Lightdash vs Holistics in 2026: pricing at 10 to 200 users, semantic layer depth, AI agents, embedding and deployment, plus other Lightdash alternatives.
Lightdash and Holistics take the same basic approach to BI: analysts define metrics in code, check them into Git, and business users explore those metrics without writing SQL. The two products differ in how far the semantic layer goes, how they price, how they embed, and where you can run them.
Holistics is an AI analytics platform that lets everyone in your organization ask questions in plain language and get reliable answers. The platform is built on a programmable semantic layer and analytics-as-code, so AI and humans reason from the same governed business logic.
We first published this comparison in 2022. Lightdash has changed a lot since then (it now has AI agents, custom charts, period-over-period comparisons, embedding and a standalone YAML mode that doesn't need dbt), so we rewrote it in November 2026 with facts checked against Lightdash's pricing page and documentation.
The short answer
- Choose Lightdash if your team already runs dbt, wants an open source (MIT) tool you can self-host for free, or has enough users that a flat fee with unlimited seats costs less than per-user pricing.
- Choose Holistics if you need metrics the semantic layer can't express in Lightdash (nested aggregations, cross-grain ratios, custom window logic), a GUI as well as code for modeling, or embedded analytics priced by end-user tier rather than by chart load.
Lightdash vs Holistics at a glance
| Aspect | Lightdash | Holistics |
|---|---|---|
| License | Open source (MIT), with enterprise code in a separate ee directory | Commercial |
| Deployment | Self-hosted (free), Lightdash Cloud (US or EU), or on-premise on Enterprise | Cloud only, with US, EU and APAC data centers |
| Paid pricing | Cloud Pro from $3,000/month with unlimited users | Entry $800/month billed annually for 10 users, then $12.50/user/month |
| Modeling language | YAML in your dbt project, or standalone Lightdash YAML | AML for models and AQL for metrics |
| Modeling interface | Code only | Code and GUI |
| Requires dbt | No (dbt Core, dbt Cloud and dbt Enterprise are all supported) | No (dbt integration available) |
| Git workflow | Content as code for charts and dashboards, plus Git sync | Native two-way Git with branching, pull requests and CI/CD |
| Period-over-period | Period comparison in the explorer and through AI agents | 1-click in the UI and composable in AQL metric definitions |
| Nested aggregations | Not supported in the semantic layer | Supported in AQL |
| Custom charts | Chart library, Chart Studio and Vega-Lite | Vega-Lite |
| AI | AI agents and MCP on Pro and Enterprise, billed by usage | AI assistant and MCP server that generate AQL against the semantic layer |
| Embedding | Add-on: iframe or React SDK, priced per embedded chart load | Separate embedded plans from $3,000/month billed annually |
| Scheduled delivery | Email, Slack and Microsoft Teams | Email and Slack |
| Compliance | SOC 2 report; HIPAA BAA for $490/month on Pro | SOC 2 Type II, GDPR; HIPAA BAA available as an add-on |
Pricing compared
Lightdash and Holistics price in opposite ways, so the cheaper option depends on your user count.
Lightdash has three options (Lightdash pricing):
- Open source: free to self-host, with community Slack support only.
- Cloud Pro: from $3,000 per month with unlimited users and unlimited visualizations. AI agents, MCP and data apps are billed by usage on top, and embedding is a paid add-on. A 21-day free trial needs no payment details.
- Enterprise: custom pricing, adding SSO with SAML, SCIM, custom roles, on-premise deployment and SLA-backed support.
Holistics prices per user on its main plans:
- Entry: $800 per month billed annually ($960 monthly), including the first 10 users. Each additional user costs $12.50 per month billed annually.
- Standard: $1,000 per month billed annually ($1,200 monthly), including the first 10 users, with more features such as custom charts and RBAC.
- Embedded analytics: separate flat plans starting at $3,000 per month billed annually for up to 5,000 embedded end users.
Here is what the base platform costs at different team sizes, before AI usage or add-ons:
| Users | Lightdash Cloud Pro | Holistics Entry (annual) |
|---|---|---|
| 10 | $3,000/month | $800/month |
| 50 | $3,000/month | $1,300/month |
| 100 | $3,000/month | $1,925/month |
| 200 | $3,000/month | $3,175/month |
At these list prices, Holistics Entry costs less up to about 186 users, and Lightdash Cloud Pro costs less above that. Self-hosted Lightdash costs nothing in license fees, but you pay for the servers and the engineering time to run it.
Semantic layer
Both tools put a semantic layer between the warehouse and the people asking questions. The difference is in what the layer can express.
Lightdash, with metrics read from a dbt project. Source: Lightdash on GitHub.
Lightdash reads dimensions and metrics from YAML, either inside your dbt project or in standalone Lightdash YAML pointed at your warehouse. It has three kinds of metrics (Lightdash metrics reference):
- Aggregate metrics (
sum,count_distinctand so on) can reference only dimensions, and wrapping one aggregate inside another causes an error. - Non-aggregate metrics (
number,booleanand so on) combine aggregate metrics, for example a ratio of two sums. - Post calculation metrics cover a fixed set of window operations:
percent_of_previous,percent_of_totalandrunning_total. They don't accept customOVER (...)clauses.
This model works well for sums, counts and ratios. Metrics like "average orders per customer by category" or a ratio against a total at a different grain usually need a derived model in dbt first.
Holistics data models
Holistics splits the layer into AML (Analytics Modeling Language) for models, relationships and dimensions, and AQL (Analytics Query Language) for metrics. AQL metrics can reference other metrics, nest aggregations, and change grain with functions like of_all(), so the definitions above stay in the semantic layer instead of moving into dbt. Analysts can write AML in the built-in IDE or work in the GUI. (For a deeper look at how semantic layers differ, see our guide to semantic layer tools.)
Self-service exploration and dashboards
Both tools let business users pick metrics and dimensions, filter, drill and build charts without SQL.
- Lightdash adds a metrics catalog, date zoom on dashboards, custom charts through its chart library and Chart Studio, Google Sheets sync, a SQL runner and pre-aggregates for faster dashboards. Period comparisons are added from the results table or requested from an AI agent (period-over-period docs).
- Holistics offers 1-click period-over-period, percent of total and moving calculations in the report builder, canvas dashboards for free-form layouts, Vega-Lite custom charts, and row-level permissions.
Both tools send scheduled reports by email and Slack, and Lightdash also supports Microsoft Teams.
AI and agents
Lightdash treats AI agents as a core feature on Cloud Pro and Enterprise. Agents answer questions from the semantic layer, run multi-step "deep research" investigations, build charts and dashboards, and can be embedded in your own product. An MCP server exposes the same metrics to tools like Claude and Cursor. Lightdash bills agents, MCP and data apps by usage (Lightdash AI agents docs).
Holistics generates AQL rather than SQL when someone asks a question in natural language, so every AI answer runs through the governed metric definitions. Holistics also provides an MCP server for external AI clients. (For more on this pattern, see our guide to AI agents for data analysis.)
Embedded analytics
Lightdash embeds charts, dashboards, AI agents, data apps and the metrics catalog through an iframe with a JWT-scoped URL or a React SDK. Embedding is an add-on for Cloud Pro or Enterprise, priced per embedded chart load: the first 1,000 loads are free, then $0.05 per load, or $790 per month for 100,000 loads per Embed Worker. A dashboard with four charts counts as four loads (Lightdash embedding docs).
Holistics embeds dashboards through an iframe with signed tokens, row-level security, dynamic data sources and white-labeling. Embedded plans are flat and sized by end users (from $3,000 per month billed annually for up to 5,000 end users), so query volume doesn't change the bill. (For other options, see our list of embedded analytics tools.)
Other Lightdash alternatives
If neither tool fits, these are the other tools teams most often compare with Lightdash:
- Looker: the original code-first semantic layer (LookML), with enterprise pricing. (For costs, see our Looker pricing guide.)
- Omni: a BI tool with a shared semantic model and a workbook-style interface for analysts. (For a side-by-side view, see Omni vs Holistics.)
- Metabase: open source (AGPL) and quick to set up, with a lighter semantic layer. Metabase Cloud starts at $100 per month for 5 users.
- Apache Superset: open source (Apache 2.0) with a SQL editor and more than 40 chart types. Preset, the managed version, is free for up to 5 users.
(For more open source options, see our guide to open source BI tools.)
FAQ
Is Lightdash free?
Yes, the open source version of Lightdash is free to self-host under the MIT license. Lightdash Cloud has no free plan, but it offers a 21-day free trial before Cloud Pro, which starts at $3,000 per month.
Does Lightdash require dbt?
No. Lightdash can read metrics from a dbt project (dbt Core, dbt Cloud or dbt Enterprise) or from standalone Lightdash YAML pointed at your warehouse.
Is Lightdash a Looker alternative?
Yes. Lightdash follows Looker's approach of defining metrics in code and exploring them through a UI, at a lower price than Looker for most teams. Holistics follows the same approach, with a more expressive metric language and a GUI for modeling.
Which is cheaper, Lightdash or Holistics?
At list prices, Holistics Entry costs less than Lightdash Cloud Pro for teams of up to about 186 users, and Lightdash Cloud Pro costs less for larger teams. Self-hosted Lightdash has no license cost. AI usage and embedding are priced separately in both products.