The short version
Take DP-800 if your daily work is SQL databases and the applications on top of them, and you want to add vector search and RAG. Take DP-700 if your daily work is moving and transforming data in Microsoft Fabric. If you do both jobs, pick the one closest to your current work first, then consider the second. Neither exam lists the other as a prerequisite, so the order is your choice.
Side by side
| Item | DP-800 | DP-700 |
|---|---|---|
| Full name | Developing AI-Enabled Database Solutions | Implementing Data Engineering Solutions Using Microsoft Fabric |
| Credential | Microsoft Certified: SQL AI Developer Associate | Microsoft Certified: Fabric Data Engineer Associate |
| Level | Intermediate | Intermediate |
| Main audience | Database and application developers | Data engineers |
| Platforms | SQL Server, Azure SQL, SQL databases in Fabric | Microsoft Fabric |
| Core languages | T-SQL | SQL, PySpark, and KQL |
| Time as published | 120 minutes | 100 minutes |
| Passing score | 700 of 1,000 | 700 of 1,000 |
| Delivery | Proctored, Pearson VUE | Proctored, Pearson VUE |
| Languages | 10 | 10 |
| Domains | 3: 35-40%, 35-40%, 25-30% | 3: 30-35% each |
| Prerequisite exam listed | None | None |
| Renewal | Yearly, free assessment | Yearly, free assessment |
Both skills lists are dated October 19, 2026. Microsoft does not publish the number of questions for either exam.
What each exam covers
DP-800 has three domains. The first is designing and developing database solutions: tables and indexes, specialized table types, JSON, programmability objects, advanced T-SQL, and AI-assisted tools such as GitHub Copilot and MCP servers. The second is securing, optimizing, and deploying: encryption, masking, Row-Level Security, execution plans, Query Store, CI/CD with SQL Database Projects, Data API builder, and change handling. The third is AI: external models, embeddings, vector and hybrid search, and RAG.
DP-700 also has three domains, each at 30-35%. The first is implementing and managing an analytics solution: workspace settings, version control and deployment pipelines, security and governance, and orchestration. The second is ingesting and transforming data: loading patterns, batch work with shortcuts and mirroring, and streaming with Eventstream, Spark structured streaming, and KQL. The third is monitoring and optimizing: monitoring items, resolving errors, and tuning performance.
Where they overlap
- T-SQL. Both expect you to read and write SQL. DP-700 uses it for transformations in warehouses and endpoints, DP-800 for deep database work.
- Security ideas. Row-level access, dynamic data masking, and permissions appear in both, in different settings.
- Lifecycle and delivery. Both cover source control and deployment, one through SQL Database Projects and one through Fabric deployment pipelines.
- Change handling. DP-800 asks about CDC and change tracking for keeping data and embeddings current. DP-700 asks about incremental loads and handling late or duplicate data.
- Performance tuning and monitoring appear in both.
Where they differ
- Center of gravity. DP-800 is database-first and application-facing. DP-700 is pipeline-first and analytics-facing.
- AI. DP-800 devotes 25-30% to AI features inside the database. DP-700 does not have an AI domain.
- Streaming and Spark. DP-700 includes streaming engines, PySpark, and KQL. DP-800 does not.
- Semi-structured data. DP-800 emphasizes JSON functions and columns inside SQL.
- Time. DP-700 has a shorter published time, 100 minutes against 120 minutes.
How to choose
- Look at your last six months of work. Mostly stored procedures, schemas, and query tuning: DP-800. Mostly pipelines, lakehouses, notebooks, and event streams: DP-700.
- Look at your next six months. Asked to add semantic search to an application database? DP-800. Asked to build a Fabric platform? DP-700.
- Check your languages. Comfortable with T-SQL but new to PySpark and KQL? DP-800 is closer. The reverse? DP-700.
- Consider both if you work across the line, for example an engineer who feeds Fabric from SQL sources and also serves vector search to an app. In that case, the overlap makes the second exam easier, though neither replaces study of the new areas.
There is no rule that you must take them in order. Taking the one that matches your daily work first usually gives you the best chance of passing, because you start from experience.
A note on versions
Both English skills lists changed on October 19, 2026, with minor changes: DP-800 in four areas, DP-700 in two. Localized versions follow about eight weeks later. Check the version offered in your language when you book.
More guides for DP-800
- DP-800 Study Plan: 4-Week and 8-Week
- DP-800 Domains Explained
- DP-800 Sample Questions With Answers
- Is DP-800 Hard? An Honest Difficulty Check
- DP-800 Prerequisites and Required Skills
- DP-800 Renewal and Retakes Explained
- DP-800 Cheat Sheet: Key Facts to Review
Next step
Read the overview of each exam: DP-800 and DP-700. Then test yourself with the DP-800 practice tests or the DP-700 practice tests. For the format details of DP-800, see the exam format page, and for its value check, is DP-800 worth it.
Frequently asked questions
Is DP-800 harder than DP-700?
They test different skills, so difficulty depends on your background. Microsoft publishes no pass rates for either.
Do I need DP-700 before DP-800?
No. Microsoft lists no prerequisite exam for either.
Which has more AI content?
DP-800. Its third domain is about AI capabilities in database solutions.