Can the Supabase server run MongoDB aggregation pipelines?

No. Supabase is backed by PostgreSQL, and execute_sql runs standard SQL. MongoDB's aggregation pipeline syntax (used by the aggregate and aggregate-db tools) is specific to MongoDB and is not available through the Supabase server.

Does either server support edge functions or serverless compute?

Supabase does: deploy_edge_function, get_edge_function, and list_edge_functions let an agent manage Supabase Edge Functions. MongoDB's server has no equivalent; it focuses on the data and Atlas infrastructure tier only.

Supabase vs MongoDB

Supabase and MongoDB both offer large, capable MCP servers, but the databases underneath are quite different: Supabase wraps PostgreSQL (relational, table-and-row, strongly typed) while MongoDB is a document store (flexible JSON-like collections, schema-optional). The Supabase server's tools span SQL execution (execute_sql), schema migrations (apply_migration, list_migrations, list_tables, list_extensions), project and organization management (create_project, pause_project, restore_project, list_projects, list_organizations), development branching (create_branch, merge_branch, rebase_branch, reset_branch), edge functions (deploy_edge_function, get_edge_function, list_edge_functions), storage (list_storage_buckets, get_storage_config, update_storage_config), logs and advisors (get_logs, get_advisors), and developer helpers like generate_typescript_types and search_docs. MongoDB's server covers document data operations (find, aggregate, aggregate-db, count, export, insert-many, update-many, delete-many), collection/index lifecycle (create-collection, create-index, collection-schema, collection-indexes, drop-collection), diagnostics (mongodb-logs, db-stats, explain), and a full Atlas control plane (atlas-create-cluster, atlas-create-free-cluster, atlas-create-db-user, atlas-create-access-list, atlas-get-performance-advisor, atlas-streams-build, atlas-streams-manage).

How they compare

Dimension	Supabase	MongoDB
Data model	Relational/PostgreSQL: tables, rows, and typed columns with foreign keys, constraints, and a rich extension ecosystem. apply_migration and execute_sql operate on SQL DDL and DML.	Document/NoSQL: schema-optional JSON-like documents in collections. find and aggregate query documents using MongoDB's own filter and pipeline syntax, not SQL.
Platform breadth	A full-stack BaaS platform: besides the database it covers edge functions via deploy_edge_function, file storage via list_storage_buckets and update_storage_config, a hosted remote endpoint with OAuth, and search_docs for documentation lookup.	A database and cloud platform: focused on the data tier plus Atlas cluster administration (atlas-create-cluster, atlas-upgrade-cluster, atlas-create-db-user, atlas-create-access-list). No file storage, no serverless functions.
Development workflow tools	create_branch, merge_branch, rebase_branch, and reset_branch give an agent branching semantics for database migrations, mirroring a Git-like workflow for schema changes.	No equivalent branching concept. Schema changes happen via direct collection and index operations (create-collection, create-index, drop-index); there is no tracked migration layer.
Diagnostics and observability	get_logs covers API, Postgres, auth, storage, realtime, and edge function logs by service type; get_advisors surfaces security and performance notices.	mongodb-logs returns recent mongod events, db-stats gives database-level usage statistics, and explain returns query optimizer plans. atlas-get-performance-advisor adds index and schema recommendations for Atlas clusters.
Authentication and setup	Requires a SUPABASE_ACCESS_TOKEN personal access token plus a --project-ref, or OAuth login via the hosted mcp.supabase.com endpoint.	Requires a MDB_MCP_CONNECTION_STRING for data tools; Atlas administration additionally needs MDB_MCP_API_CLIENT_ID and MDB_MCP_API_CLIENT_SECRET for the service-account key pair.

Verdict

Choose Supabase when your application runs on the Supabase BaaS platform and you want the agent to participate across the stack: applying migrations, deploying edge functions, managing storage, and debugging via logs, all with Postgres as the underlying database. Choose MongoDB when your data model favors flexible documents over rigid tables, or when you are already on MongoDB or Atlas and want full cluster administration alongside document queries and aggregation pipelines. The servers are large and capable in different directions; neither is a drop-in substitute for the other.

FAQ

Can the Supabase server run MongoDB aggregation pipelines?: No. Supabase is backed by PostgreSQL, and execute_sql runs standard SQL. MongoDB's aggregation pipeline syntax (used by the aggregate and aggregate-db tools) is specific to MongoDB and is not available through the Supabase server.
Does either server support edge functions or serverless compute?: Supabase does: deploy_edge_function, get_edge_function, and list_edge_functions let an agent manage Supabase Edge Functions. MongoDB's server has no equivalent; it focuses on the data and Atlas infrastructure tier only.