Simpler sidebar. Import files and folders with ease and make folders when needed. Just drag your files on browser not just on a single button.

Try it out at https://datakit.page/

Changelog link:

We heard from our Brazilian friends that PT language would be super nice to have. Here we go!

Try it out at https://datakit.page/

Changelog link:

if you want to persist the result of your ran queries, you can easily save them with "Save as Table" shown on top of query results in the header.

How it works:

Try it out at https://datakit.page/

Changelog link:

From now on, each file you bring to datakit, gets its own state of query, assistant and preview. So not just one query, viz, assistant at a time, as many as number of the files you have.

Try it out at https://datakit.page/

Having an instant look at distribution, uniq, nulls and more from your column.

Try it out at https://datakit.page/

From now on you could simply connect your postgres databases to datakit. Preview, query and ask AI assistant to write a query for the tables you have questions about.

Try it out at https://datakit.page

Ever had to re-import the same CSV files every time you opened your browser? Or lost track of which datasets belonged to which analysis project?

Workspaces eliminates these friction points.

This is just the beginning of making data analysis more seamless and less repetitive - because your time should be spent on insights, not file management.

Try it out at https://datakit.page and let us know what you think!

DataKit now supports Ollama integration! Run your favourite local AI models (Llama 3, Mistral AI, DeepSeek r1, Microsoft Phi-4, Qwen3, OpenAI gpt-oss and more) directly in your data workflows.

Complete privacy - your data never leaves your machine, no API costs or rate limits and full control over your AI infrastructure.

All you need to do is run Ollama (download it from: https://ollama.com) and open up https://datakit.page.

Introducing "" in DataKit.

From now on you can run your python notebooks in https://datakit.page.

You can bridge to your DuckDB from your notebook and write the sql, make plots, etc.

Additionally, transformers from Hugging Face are loaded by default in the environment.

For now the supported models are from:

These are all in your browser, there's no server (yet).

Watch a demo from here:

Check it out :)

The models are

claude-3-5-sonnet-20241022

claude-3-5-haiku-20241022

On coming months, the selection of desired models will get broader