What are the advantages of wrapping a react application using <ApolloProvider> as compared to make mutations and queries directly using javascript

I'm new to Apollo Client React. I was reading the docs and i found this <ApolloProvider> wrapper which seems scary to me. My question is why do we need <ApolloProvider> if we can make queries directly and mutations using vanilla js.. I mean why don't we just make queries for example like this:

...
const client = new ApolloClient({
  uri: "http://localhost:3001/graphql",
  cache: new InMemoryCache(),
});

...
  client
    .query({
      query: gql`
        {
          books {
            name
            id
            author {
              name
            }
          }
        }
      `,
    })
    .then((result) => console.log(result.data));

and then store the results in a state. I'm not actually asking Why? per say. I'm looking for the advantages of this approach of a wrapper <ApolloProvider> over making queries and store them in redux-state. Making mutations directly by calling client.mutate(...) without bothering using the useQuery() apollo hooks.