[Answer]-Django cache query result for templates and views

You could modify the solution proposed here to use per-request locmemcache instances, instead of a global cache per thread.

You could then use the middleware’s process_response function to then cleanup the local cache.