Quickstart

The fastest way to experience Vectara immediately is by querying our pre-built sample data.

1. Navigate to the Corpora page.
2. Click the drop-down on the Create corpus button, and select Try a sample corpus.
   The sample data populates the corpus and takes you to the Query tab of the samples corpus.
   tip

   Want to use your own data? Read the Data ingestion guide to create a new corpus and upload your documents (PDFs, TXT, etc.).

3. Type a query like What's inside a black hole? and press Enter.
   If you're using your own data, type a query relevant to the data that you uploaded.
   
   The score value is our confidence in the match. The text under each result is a snippet from the source document.
4. You can experiment with different retrieval and generation options like hybrid search, reranking, and summarization.

Search for answers with the Query API

Tip

Use our Query APIs directly or use our Vectara Python SDK if you're working on Python applications.

Query an existing corpus and get AI-generated answers with context. In this example, you query the black-holes-sample-data corpus.

EXAMPLE API QUERY

Code example with bash syntax.bash

1

Let’s take a closer look at the first response:

EXAMPLE JSON RESPONSE

Code example with json syntax.json

1

The result answers the question and returns additional details about the query, such as the language, section, and offset.