Ddbt MCP (official)

Tested PyPI dbt-mcp 2.10.1 (Apache-2.0; repo dbt-labs/dbt-mcp, 620 GitHub stars when checked; server reports dbt 1.30.0). Stdio, 39 tools with no env set (product docs, Semantic Layer, Discovery, and Admin API job tools incl. trigger, cancel and retry job run). Real calls: search_product_docs for 'incremental models' and get_product_doc_pages for the incremental-models page returned content. list_metrics answered that DBT_HOST is required, so platform tools need a dbt host and token. Core tools not tested beyond docs; the job-run tools were not called. dbt's docs give 5,000 requests per minute per IP for the remote dbt MCP only, not this local server, so no rate limit is shown.
Tool-call test (2026-11-11, direct MCP stdio client, no LLM): partial. 2 of 3 read calls passed; Semantic Layer and platform tools need a dbt host and token. Evidence: local test run, Nov 11, 2026.

Transport (tested 2026-11-11): stdio

Score 90/100 · Tools 39 · Token cost approx. 18,871 tokens · Compare · Last tested 2026-11-11 · Version tested 2.10.1 · Package dbt-mcp · Repo dbt-labs/dbt-mcp

Summary: dbt MCP (official) completed an MCP handshake without credentials and exposed 39 tools.

Evidence summary

Handshake

pass

Tools listed

39

Context cost

approx. 18871 tokens, cl100k estimate (39 tools, measured 2026-11-11, version 2.10.1; an estimate) ranking · add up your servers' context cost

API key

Connects without a key; tool calls need one

Tested

2026-11-11

Test results

Local-model readiness

Token cost (estimate): approx. 18871 tokens, cl100k estimate for 39 tools · about 230% of an 8k context. This is an estimate from our testing (counted with cl100k_base).

Small-model friendly (schema heuristic, not a model result): unknown.

What we tested and what we didn't

Methodology: how we test. A high score means it installs and responds, not that giving it access to your data carries no risk.

{
  "name": "dbt-mcp",
  "version": "2.10.1",
  "registry": "pypi",
  "status": "working",
  "tested": "2026-11-11",
  "tools_count": 39,
  "deprecated": false,
  "archived": false,
  "tests": {
    "install": [
      "pass",
      20,
      "`uv pip install dbt-mcp` (Python venv, Linux) succeeded in 1s."
    ],
    "handshake": [
      "pass",
      30,
      "MCP `initialize` + `tools/list` over stdio succeeded with no credentials; 39 tools listed in 2.9s."
    ],
    "maintenance": [
      "pass",
      25,
      "Repo pushed 1 days before test; 620 GitHub stars."
    ],
    "docs": [
      "partial",
      5,
      "README over 3,000 chars. Check of the package's own README (PyPI long_description (package README))."
    ],
    "license": [
      "pass",
      10,
      "Apache Software License (OSI-approved)."
    ]
  }
}

Install

uvx dbt-mcpCopy

Pin the version we tested

uvx dbt-mcp==2.10.1Copy

pip install dbt-mcp==2.10.1Copy

tested 2026-11-11. Newer versions may differ and were not tested by us.

{
  "mcpServers": {
    "dbt-mcp": {
      "command": "uvx",
      "args": [
        "dbt-mcp==2.10.1"
      ]
    }
  }
}

Tools listed

cancel_job_run, get_all_macros, get_all_models, get_all_sources, get_dimension_values, get_dimensions, get_entities, get_exposure_details, get_exposures, get_job_details, get_job_run_artifacts, get_job_run_details, get_job_run_error, get_lineage, get_macro_details, get_mart_models, get_metrics_compiled_sql, get_model_children, get_model_details, get_model_health, get_model_parents, get_model_performance, get_node_details, get_product_doc_pages, get_seed_details, get_semantic_model_details, get_snapshot_details, get_source_details, get_test_details, list_job_run_artifacts, list_jobs, list_jobs_runs, list_metrics, list_projects, list_saved_queries, query_metrics, retry_job_run, search_product_docs, trigger_job_run

Tool history: captured Nov 11, 2026. All tool history.

Quick config

{
  "mcpServers": {
    "dbt-mcp": {
      "command": "uvx",
      "args": [
        "dbt-mcp"
      ]
    }
  }
}

Add to config Opens the config generator with this server selected.

{
  "mcpServers": {
    "dbt-mcp": {
      "command": "uvx",
      "args": [
        "dbt-mcp"
      ]
    }
  }
}

Related: Best data and analytics MCP servers

FAQ

Is dbt MCP deprecated?

No. dbt-mcp was not deprecated or archived when we last tested it on 2026-11-11.

What replaces dbt MCP?

Nothing; dbt MCP (official) is not deprecated as of 2026-11-11.

Does dbt MCP need an API key?

It connects and lists 39 tools without a key, but its tool calls need one.

Embed the badge

Every tested server gets a free embeddable badge. It shows the current score and test date and expires after 90 days. Testing is free; payment never changes a score.

Copy HTML

Copy Markdown Claim this listing · Report an issue