Launching today
FluentDB
The AI database client for Mac
239 followers
The AI database client for Mac
239 followers
FluentDB is an AI-powered database client for macOS. It supports PostgreSQL, MySQL, SQLite, SQL Server, with more coming soon. Built for speed, simplicity, and security, it works with Anthropic and OpenAI using your own API keys or Claude Code/Codex subscriptions, and with local models via Ollama. Your data never leaves your machine without your approval: models see your schema, not your data, and every AI-generated SQL query requires approval by default.
FluentDB
FluentDB
@thys_beesman Thanks for your comment!
> UPDATE without a WHERE clause or a query that touches ten thousand rows
Yes, I agree with you. By default, even if you accept a query, when it's a write query, we prompt for another confirmation to make sure you didn't just press Enter by reflex. At this moment, we don't show the potential number of impacted rows; however, it's a great idea, and I just added this to our roadmap :)
Another thing is that we made it super easy to kill a running query (server-side) in one click. Of course it wouldn't prevent big trouble in case of a wrong write, but at least if it's a long running read that can lock your database, you can just kill the process in one click. All our guardrails are explained here.
> Is there any way to selectively hide or rename specific tables and columns from what gets sent to the AI,
Yes, I understand this concern. For now, we don't have this kind of system; however, it's easy to put in place, so I will add it to our roadmap. Something to note is that if the security is the main concern, you can still use a local model using Ollama so it would completely remove every risk of data leak since you control the model.
Thank you for your feedback it's really useful. Feel free to reach out anytime if you give it a try and want to help improve the tool :)
a native Mac db client that isn't electron is already a plus in my book. curious how the AI part handles write queries specifically - does it always show you the generated SQL before running anything against the actual database, or can it execute directly if you ask it to "clean up duplicate rows" type stuff without a confirmation step?
FluentDB
@omri_ben_shoham1 Thanks for your comment!
By default, the AI will always only write queries and ask you to confirm before executing them. Even when you run them, we don't share the result with the AI.
However, yes, you can turn on two permissions:
- Auto run: allows the AI to run the queries it wants
- Read mode: allows the AI to view the results
I explain this in detail in this video :)
The Ollama path is the obvious choice when the database has anything sensitive — inference runs locally, nothing leaves. One thing I want to understand: when you use a local model, what actually goes into the AI context? Is it the schema and the current query only, or do prior query results get included to give better suggestions? I'm wondering whether the opt-in read mode applies to local models the same way it does to cloud ones, or if the local boundary means results are always excluded.
FluentDB
@hi_i_am_mimo The permissions are managed the same way even if it's a local ollama connection, because some people use ollama as a local server but proxy it through an external provider ; so we cannot just assume that ollama is safe by default.
However, you can give full access to the AI in that case, and then the AI will have the full context filled with all results in the conversations.
Nice one, native Swift instead of another Electron app is already a win for me. The schema only approach for the AI is the right call. When you say write queries need extra confirmation, can I lock it to read only mode entirely for certain connections?
FluentDB
@adamkamaneh Thanks for your comment! Yes, absolutely. When you create a connection, you can choose a read-only mode, and in that case, no write queries are allowed or suggested by the LLM.
Congrats team! I have been burned before by an AI tool that wanted way too much access to my data, so the approval by default thing is really reassuring. Curious about the Ollama support, does the experience feel as smooth with local models or is it noticeably better with the bigger providers?
FluentDB
@doganakbulut thanks for the support!
To answer your question, so far I have tried Ollama with the following models: qwen3-coder, gemma4, qwen4.6.
Qwen3-coder is the one that gave me the best result in terms of speed and relevance.
However, I have to admit that I personally use my Claude Code subscription with Haiku most of the time because it's just much faster and more relevant from my experiments.
@kevinpiacentini that confirm-before-execute default plus not sharing results back unless you opt into read mode is exactly the design I was hoping for, opt-in auto-run and opt-in read mode as two separate toggles is a nice way to let people dial in their own risk tolerance instead of one blanket switch. will check out the video, thanks for the detailed answer.
FluentDB
@omri_ben_shoham1 thank you for the feedback :)