Overview

Florentine connects to MongoDB Atlas, MongoDB Community Edition, and MySQL through connection strings. It analyzes table and collection schemas to generate SQL or MongoDB aggregation pipelines suited to the connected database. That focus can serve teams working within those systems, but it is not a fit for stacks that require broader database support.

Key features

Query generation and review

Users can receive a generated query, its results, or a natural-language response. Read-only queries can run automatically or wait for review. This gives teams a choice between faster self-service and an extra checkpoint before reads execute, though the control is specifically for read-only queries rather than general approval of database operations.

LLM, agent, and search options

Florentine supports OpenAI, Deepseek, Google, and Anthropic with bring-your-own-key access, so teams can use those providers while supplying their own credentials. Its API and MCP server have guides for Claude, ChatGPT, and Cursor, making it relevant to teams connecting database queries to existing tools. MongoDB vector search uses automated embeddings; MySQL semantic re-ranking uses a self-hosted embedding model, a distinction teams should weigh when planning their search setup.

Isolation and data handling

Mandatory user-based query filters support tenant separation without depending on the LLM to apply them. The privacy policy says the application and MongoDB Atlas databases are hosted in EU Frankfurt regions, with standard contractual clauses for possible US transfers. It also names TLS, AES-256 encryption at rest, MFA, and role-based access controls. Customers can request a DPA by email.

Pricing

The Free plan costs 0.00 EUR per free, billed forever. It includes one database, five active collections or tables, 500 requests per month, MongoDB and MySQL support, and the MCP server and API. It is a useful way to try the core integration pattern, but the request cap and small schema allowance make it better suited to evaluation or light workloads than broad team use.

Pro costs 39.00 EUR per month, billed per month. It keeps the one-database limit while increasing capacity to 25 active collections or tables and 2,500 requests per month. This is the more practical tier for a single-database team with recurring use; it does not expand the number of databases.

Enterprise costs 149.00 EUR per month, billed per month, and supports five databases, 50 active collections or tables, and 10,000 requests per month. It is the relevant step up for teams needing multiple database connections and higher quotas. All three plans include MongoDB and MySQL, API, and MCP access. Plan terms do not state seat limits, so the published quotas are the clearest basis for comparing capacity.

Platforms

Florentine is a cloud product with web access, an API, and an MCP server. The supported database connections are MongoDB and MySQL. Teams needing self-hosted deployment or other database families should consider alternatives.

Who it's for

Florentine suits support teams, business analysts, internal search builders, recommendation-engine teams, and other groups that want natural-language access to MongoDB or MySQL. Its review option and user-based filters matter most where database access needs oversight or tenant separation. It is less compelling for organizations that need self-hosting, a wider database catalog, or quotas beyond the plans' stated limits.

Pros and cons

Pros: Schema-aware generation for MongoDB and MySQL avoids requiring users to compose queries manually.
Pros: Users can inspect queries or results and control automatic execution of read-only queries.
Pros: API, MCP, multiple BYO-key LLM providers, and user-based isolation filters support integration into existing workflows.
Cons: Database support is limited to MongoDB and MySQL, and deployment is cloud-only.
Cons: Free is capped at one database, five active collections or tables, and 500 monthly requests; Pro still allows only one database.
Cons: The monthly plans specify request and schema quotas but no seat limits.

Alternatives

Compare AI database assistants if you want to survey the category. Outerbase AI is worth considering when its macOS, Windows, or self-hosted options matter; its Free plan includes up to five users but only 10 EZQL queries per month and is limited to transactional databases. Wren AI is a better direction for developers seeking a free open-source context engine used through CLI and MCP without a UI, or a self-hosted option.

Kyligence is another paid alternative with API, Linux, self-hosted, and web platforms. AI for Database may suit teams wanting self-hosting; its Pro plan is 20.00 USD per month with $20 credits per payment, unlimited workflows, and credits that do not expire. Snowflake Feature Store is a paid, consumption-based option centered on Snowpark, data sharing, optimized storage, and Time Travel.

Tablix is a free alternative with API, self-hosted, and web platforms. Vanna AI offers a $50 per month Explorer plan with 20 questions per day, an API, and same-day email support. AnswerRocket uses customized pricing based on users, use cases, data sources, and services.

Verdict

Choose Florentine if your team works with MongoDB or MySQL and needs natural-language querying that can connect to existing tools while retaining review controls and user-based isolation. Its practical limitation is the tight database scope, compounded by quotas that may force an upgrade as usage grows; teams needing self-hosting or broader database coverage should look elsewhere.

	Florentine.ai	Snowflake Feature Store
Free plan	Yes	No
Free trial	Yes	No
Paid from	€39/mo	—
Open source	No	No
Platforms	api, Web	api, Web
Free plan	Yes	—
Natural-language queries	Yes	—
Deployment	cloud	—
Supported databases	MongoDB, MySQL	—

Florentine.ai vs Snowflake Feature Store

Florentine.ai

Snowflake Feature Store