Florentine.ai

Florentine.ai lets users query MongoDB and MySQL data in natural language instead of writing queries themselves. It connects to MongoDB Atlas, MongoDB Community Edition, and MySQL through connection strings, then analyzes schemas to inform query generation. Depending on the database, it creates MongoDB aggregation pipelines or SQL queries. Users can choose raw queries, query results, or natural-language answers, and decide whether read-only queries run automatically or wait for review. Search features include MongoDB vector search with automated embeddings and semantic re-ranking for MySQL using a self-hosted embedding model. Florentine supports OpenAI, Deepseek, Google, and Anthropic through a bring-your-own-key approach. Its MCP server and API have guides for Claude, ChatGPT, and Cursor. The cloud service is described for support teams, business analysts, internal search tools, recommendation engines, and other database users. The Free plan covers one database, five active collections or tables, and 500 requests per month. Pro costs €39.00 per month and allows 25 active collections or tables and 2,500 monthly requests. Enterprise costs €149.00 per month and supports five databases, 50 active collections or tables, and 10,000 monthly requests.

Florentine.ai combines natural-language querying with options for query review and multiple response formats. Compare the Free plan's database, collection, and request limits with the higher Pro and Enterprise allowances.

Florentine.ai suits teams that want to explore MongoDB or MySQL data using natural-language requests. The listed use cases include support, business analysis, internal search, and recommendation tools.

Florentine.ai is a cloud service that turns plain-language questions into MongoDB or MySQL queries. It fits support teams, analysts, and developers building database access into internal tools, especially when query review and agent integrations matter. Its narrow database coverage is the main boundary.

Overview

Florentine connects to MongoDB Atlas, MongoDB Community Edition, and MySQL through connection strings. It analyzes collection and table schemas, then generates MongoDB aggregation pipelines or SQL that match each system's capabilities. That schema-aware approach is more useful than a generic text-to-query layer when your data structure changes across collections or tables, but it does not help organizations whose primary data sits outside MongoDB and MySQL.

The service is cloud-deployed and headquartered in Deutsch Evern, Germany. Its application and MongoDB Atlas databases are hosted in EU Frankfurt regions, with standard contractual clauses covering possible transfers to the US.

Key features

Natural-language querying and review

Users can request the raw generated query, the resulting data, or a natural-language answer. Read-only queries may run automatically or be held for review. Automatic execution suits routine self-service questions; review mode gives analysts and support staff a chance to inspect the generated operation before it runs. This control is specifically for read-only queries, so it should not be treated as a universal approval workflow for database changes.

AI and agent connections

Florentine supports OpenAI, Deepseek, Google, and Anthropic using a bring-your-own-key arrangement. That can give teams control over their chosen provider, while also making provider-key management part of deployment. An MCP server and API, with guides for Claude, ChatGPT, and Cursor, make the product suitable for embedding database questions in assistant or developer workflows rather than keeping every interaction in a standalone interface.

Search and data isolation

MongoDB vector search includes automated embeddings. For MySQL, semantic re-ranking uses a self-hosted embedding model, a useful distinction for teams that want semantic retrieval while retaining control of that model. Tenant isolation adds mandatory user-based query filters without relying on the language model to apply them, which is a meaningful safeguard for multi-user applications. Security measures include TLS, AES-256 encryption at rest, multifactor authentication, and role-based access controls. Customers can request a data-processing agreement by email.

Pricing

Florentine uses a freemium model with a permanent Free tier and paid plans from €39 per month. The Free plan costs nothing, billed forever, and includes one database, five active collections or tables, 500 requests per month, MongoDB and MySQL support, and the MCP Server and API. It is suitable for a small proof of concept or a narrowly scoped internal tool. The one-database, five-structure, and 500-request ceilings will constrain a growing team quickly.

Pro costs €39 a month, billed monthly. It keeps one database but raises the allowance to 25 active collections or tables and 2,500 requests per month, while retaining both database types and the MCP Server and API. This is the practical choice for a single production data source with moderate team traffic, but it offers no increase in database count.

Enterprise costs €149 a month, billed monthly. It supports five databases, 50 active collections or tables, and 10,000 requests per month, with the same MongoDB, MySQL, MCP Server, and API coverage. Choose it when separate databases or materially higher request volume justify the price; smaller deployments gain little from its extra capacity. A free trial is offered, but no trial duration is stated.

Platforms

Florentine is available through its web application and API. The MCP server extends access to compatible assistants and developer tools, while deployment remains cloud-based.

Who it's for

Support teams can expose database answers without asking every agent to learn query syntax. Business analysts gain a choice between inspectable queries, returned results, and prose answers. Developers can use the API or MCP server for internal search tools and recommendation engines. It is a poor fit for teams requiring databases beyond MongoDB and MySQL or a desktop-first product.

Pros and cons

• Pros: Schema analysis produces database-specific aggregation pipelines or SQL instead of one generic query format.
• Pros: Review or automatic execution for read-only requests balances oversight with speed.
• Pros: MCP, API, and four bring-your-own-key LLM providers support assistant-centric workflows.
• Pros: Mandatory tenant filters and documented encryption, MFA, and role controls address shared-data deployments.
• Cons: Support is limited to MongoDB and MySQL, excluding teams on other database systems.
• Cons: The Free tier's one database, five active collections or tables, and 500 monthly requests leave little room for expansion.
• Cons: Pro still permits only one database, so multi-database organizations must move to the €149 Enterprise tier.

Alternatives

Browse the AI Database Assistants list for a broader category comparison.

Choose ThoughtSpot instead if you need web, iOS, and Android access and a paid analytics product with annual Essentials pricing of $25 a month for 5 to 50 users and up to 25M rows. Choose DataZen for a free Linux, macOS, or Windows option with no account and a bring-your-own AI provider. DBQuill is a free, open-source Windows desktop application when a source-first local tool is preferable.

Insight O' Mate suits desktop users on Linux, macOS, or Windows who want a free tier with 20 database analyses per day, or a $29 monthly Pro plan with unlimited queries, priority queue, and advanced export. Outerbase AI is worth considering for broader platform coverage, including self-hosted and desktop options. Wren AI is an alternative with a free open-source context engine for individual developers and a free trial on its cloud offering. YourQL is a free cross-platform desktop application still described as a work in progress. DataLine is a free open-source project released under GPL-3.0.

Verdict

Florentine.ai is best for teams that already run MongoDB or MySQL and want natural-language access with schema-aware generation, reviewable read-only queries, and API or MCP integration. Choose it for those controls and isolation features; look elsewhere if you need wider database support, desktop software, or more than one database without paying for Enterprise.