mnemiq

Free plan APISelf-hostedWeb

What mnemiq is

mnemiq is an open-source text-to-SQL engine for enterprise data agents, configured and measured on a user's own database. Before a query runs, a deterministic checker verifies that proposed SQL is read-only, uses allowed objects, compiles in the source dialect, and passes EXPLAIN. Permissions apply before schema retrieval, so the model is not shown tables the caller cannot access. Answers include the SQL executed, the tables it touched, and the enrichment version used. The engine can profile columns and build descriptions, grain, glossary terms, and coded-value meanings; a human owner can certify semantic definitions. It lists PostgreSQL, SQLite, DuckDB, Oracle, Snowflake, and Databricks support, with DuckDB as the universal executor. Any OpenAI-compatible /v1 endpoint can provide the model, including local servers such as vLLM, Ollama, llama.cpp server, and LM Studio. Its MCP server exposes read-only, access-scoped db_read(question) and get_schema() tools. One process can serve a workbench and HTTP endpoints for asking, chatting, and retrieving schema. The engine is free under Apache-2.0 and runs in the user's environment. Separate commercial services are also identified around the engine.

Who it suits

mnemiq suits teams building enterprise data agents that need to query their own databases with access-scoped schema visibility. It may also fit users seeking a self-hosted engine and support for local model servers.

What it does well

Free open-source engine under Apache-2.0.
Checks proposed SQL before execution.
Permissions apply before schema retrieval.
Answers include executed SQL and touched tables.
Supports multiple databases and model endpoints.

What to know first

The engine must run in the user's environment.
Benchmark comparison reports around 2.3 LLM calls per question with repair enabled.
Managed services around the engine are commercial.

Limits

The free mnemiq engine runs in your own environment. Managed grading, drift services and the Agentic Fabriq control plane are separate commercial offerings; Fabriq lists custom Enterprise pricing, without a mnemiq-specific price.

Plans

Plan	Billed and limits	Price
mnemiq	Open-source Apache-2.0 engine · runs in your own environment	Free
Verity and Agentic Fabriq control plane	Pricing page lists custom Enterprise pricing for Fabriq; it does not state a mnemiq-specific price. · Managed grading and drift service · identity, vaulted credentials, per-group grants and audit across sources	On request

As agenticfabriq.com lists them, read 4 Oct 2026. The maker's own page has the final word.

Getting started

Visit the mnemiq page at agenticfabriq.com/mnemiq
Run the Apache-2.0 engine in your own environment
Connect a listed database and an OpenAI-compatible /v1 model endpoint
Use the workbench, HTTP endpoints or MCP tools

Verdict

mnemiq emphasizes read-only query checks, access control, and traceability while supporting multiple databases and model endpoints. Users should account for its own-environment deployment and the separate commercial services around it.

Questions

Is mnemiq free?

Yes. The mnemiq engine is listed as free under the Apache-2.0 license and runs in your own environment.

What databases does mnemiq support?

It lists PostgreSQL, SQLite, DuckDB, Oracle, Snowflake and Databricks. DuckDB is identified as the universal executor.

Can mnemiq use a local model?

Yes. Any OpenAI-compatible /v1 endpoint can serve the model, including local servers such as vLLM, Ollama, llama.cpp server and LM Studio.

How can agents access mnemiq?

Its MCP server exposes the read-only, access-scoped tools db_read(question) and get_schema(). A single process can also serve a workbench and HTTP endpoints for asking, chatting and retrieving schema.

Who maintains mnemiq?

The mnemiq page identifies Agentic Fabriq as its maintainer.

Are Verity and the Agentic Fabriq control plane included with the free engine?

They are identified as commercial services around the open-source engine. The pricing page lists custom Enterprise pricing for Fabriq, but no mnemiq-specific price.

Facts from mnemiq's own pages

Access control: Permissions are applied before schema retrieval, so the model is not shown tables the caller may not access. · agenticfabriq.com, 4 Oct 2026
Agent integration: The MCP server exposes two read-only, access-scoped tools: db_read(question) and get_schema(). · agenticfabriq.com, 4 Oct 2026
API and workbench: One process can serve a workbench and HTTP endpoints POST /v1/ask, POST /v1/chat, and GET /v1/schema. · agenticfabriq.com, 4 Oct 2026
Commercial services: The maker identifies Verity, a managed grading and drift service, and the Agentic Fabriq control plane as commercial services around the open-source engine. · agenticfabriq.com, 4 Oct 2026
Company: The mnemiq page identifies Agentic Fabriq as its maintainer; the opened maker pages did not state a headquarters or founding year. · agenticfabriq.com, 4 Oct 2026
Cost consideration: The launch article says mnemiq averaged around 2.3 LLM calls per question with repair enabled in its benchmark comparison, while the compared vendor APIs were called once. · agenticfabriq.com, 4 Oct 2026
Database support: The page lists PostgreSQL, SQLite, DuckDB, Oracle, Snowflake, and Databricks, with DuckDB as the universal executor. · agenticfabriq.com, 4 Oct 2026
LLM support: Any OpenAI-compatible /v1 endpoint can serve the model, including local servers such as vLLM, Ollama, llama.cpp server, and LM Studio. · agenticfabriq.com, 4 Oct 2026
Purpose: mnemiq is an open-source text-to-SQL engine for enterprise data agents that can be configured and measured on a user's own database. · agenticfabriq.com, 4 Oct 2026
Query checks: A deterministic decider checks that proposed SQL is read-only, uses allowed objects, compiles in the source dialect, and passes EXPLAIN before execution. · agenticfabriq.com, 4 Oct 2026
Security deployment: The maker says that using a local model server keeps every schema, question, and row inside the user's network. · agenticfabriq.com, 4 Oct 2026
Semantic enrichment: It can profile columns and build descriptions, grain, glossary terms, and coded-value meanings, with semantic definitions optionally certified by a human owner. · agenticfabriq.com, 4 Oct 2026
Traceability: Each answer includes the SQL that ran, the tables it touched, and the enrichment version used. · agenticfabriq.com, 4 Oct 2026

Where mnemiq stands on Freedom251

AI SQL Generators#3 of 19