Improved Query Autocomplete with Trie-Based Suggestions
Mongoose Studio’s query editor now uses a Trie-based autocomplete engine for lightning-fast, context-aware suggestions.
The new implementation indexes both MongoDB query operators (like $gte, $in, $elemMatch, etc.) and all schema paths from the connected model, giving instant, ranked completions as you type.
Suggestions now appear more consistently and handle partial or slightly misspelled inputs thanks to built-in fuzzy matching, dramatically improving the feel of the search and filter workflow.
Behind the scenes, the editor builds and maintains a lightweight Trie every time a model or schema changes, ensuring suggestions always stay up-to-date with the current collection.
Added Support for Local Chat Completions via OpenAI API
Mongoose Studio can now run chat completions locally using your own OpenAI API key, without relying on the Mongoose Studio mothership.
The new openAIAPIKey option allows developers to pass process.env.OPENAI_API_KEY when mounting Studio, enabling direct calls to OpenAI’s /v1/chat/completions endpoint for both chat responses and automatic thread summarization.
This makes chat completions faster, because they skip the extra network hop to the Mongoose Studio mothership.
app.use('/studio', await studio(
null,
conn,
{
apiKey: 'studio API key here',
model: 'gpt-4.1-nano', // Optionally set model. Mongoose Studio uses gpt-4.1-nano by default.
openAIAPIKey: 'sk-proj-MYOPENAIKEY'
})
);
Added Shift-Click Multi-Select and Improved Selection UX
The Models view now supports Shift-click multi-selection, making it much easier to select ranges of documents for batch operations like delete or export. When multi-select mode is active, holding Shift while clicking selects all documents between the last selected row and the newly clicked one, just like in VS Code or file explorers. The “Select” button also now toggles dynamically between Select and Cancel for clearer feedback when entering or exiting multi-select mode.