Your team stopped reading the docs months ago. Now you're the human search engine, fielding the same Slack questions, screen-sharing your DAG in meetings, tracing lineage by hand. Fix that in 30 seconds.
Find any model in seconds. Trace lineage to the column. Stop being your team's human search engine.
Unlimited viewers, $99/month. Give your whole company dbt docs. Product, finance, engineering. Everyone.
One command in CI. Zero config. Static site or hosted. MCP server included.
The Jaffle Shop project with columns expanded. Try it live.
dbt docs serve was built for small projects. Docglow generates a fast, static site that scales with yours.
Full-text search across models, sources, and macros. Type a name, get a result. Done scrolling through tree views.
Interactive lineage graph with column-level tracing. Stop spending 20 minutes following a metric through five layers of CTEs.
Test results and source freshness at a glance. See which models need attention before someone Slacks you about it.
Nine built-in MCP tools. Cursor, Claude Code, and other AI editors can explore your models, lineage, and metadata through them.
Clean React UI with dark mode. It looks like a product built in 2025, because it is.
Point it at your dbt project and go. Works with your existing manifest and catalog. You keep your data, your hosting, your workflow.
Configurable data layers group models by staging, transformation, and mart. Spot that mart model pulling from a staging table in two clicks.
See your full DAG organized by data layer. Click any model to see its dependencies, tests, and documentation. Search across your entire project from one bar.
Expand any model to see its columns. Trace a single column upstream through staging to source. Dashed lines show you the exact path.
Every column shows its type, description, upstream sources, and downstream consumers. Your PM can look up a field without asking you.
dbt Cloud charges $100+/seat. Data catalogs start at $50K/year. You pay $99/month and give every person in your company access.
Stakeholders ask data questions in plain English. They get answers. You stay focused.
Your team asks data questions in Slack and gets answers there. They stay in context. You stay out of it.
Track documentation coverage, test health, and freshness trends. Know where your project needs attention before it becomes an incident.
Publish to a custom subdomain from your CI pipeline with one command. Docs update on every merge.
MIT-licensed. The CLI generates a static site you host wherever you want. We built Docglow because we ran dbt in production for years and hated the docs experience the entire time.
See what self-serve dbt docs look like. Takes 30 seconds.