Static ReadonlyAuthorizesTrue when the action authorizes its caller itself, so a public entry point such as the
RunAction mutation may run it for an authenticated user without an MJ: Action Authorizations link. Set it only on an action that either checks an authorization the caller
must hold, or reads and writes only with the caller's own permissions. Leave it false for an
action that acts with more authority than its caller, or that reaches data or sends messages
the caller chooses but could not otherwise reach.
The engine reads the marker only from the class that runs, never from its parent classes. A subclass that overrides a marked action must declare the marker again.
Whether this action can bound what it returns to an audience (RunActionParams.Audience): everyone
besides the caller who will see its output, such as the other participants of a shared conversation.
false by default, and the engine enforces it: when a run's audience adds a reader beyond the caller,
ActionEngineServer.RunAction refuses the action with result code AUDIENCE_UNSUPPORTED, without running
it. Override to return true only when the action honours the audience in full — every record, snippet
and count it returns is something every reader may see (pass the audience to the search engine, check each
reader's entitlement to anything it scopes by, and leave out aggregates computed before that filtering).
ProtectedInternalExecutes the Perplexity call.
The action parameters containing:
search mode only:
answer mode only:
llama-3.1-sonar-* identifiers were retired by Perplexity in
February 2025 and now fail with an invalid-model errorStructured results in search mode; prose plus citations in answer mode
Executes the action with the provided parameters.
The action execution parameters including context
Promise resolving to the action result
Action that queries Perplexity. It fronts two different products, selected by
Mode, because they answer different questions and cost different money.search(default) callsPOST /search, Perplexity's raw search endpoint. It returns rankedtitle/url/snippetresults with no model in the loop: sub-second, flat-priced per query, and deterministic enough to cache. This is the mode to use when something wants sources.answercallsPOST /chat/completionsagainst the Sonar models. It returns synthesised prose plus a flat citation list, takes seconds rather than milliseconds, and bills per token. Use it when a caller genuinely wants a written answer rather than results to read.searchis the default deliberately. Every caller that omittedModebefore this parameter existed was paying Sonar generation prices, and waiting for generation latency, to obtain what was usually just a list of links — andCitationsis still populated insearchmode, so callers that only read the URL list keep working unchanged.Example