supports
fun supportsStreaming(llmOperations: LlmOperations, llmOptions: <Error class: unknown class>): Boolean
Tests whether the LLM resolved from the given operations and options supports streaming.
Only a definitive yes is stored, and only when modelNameCacheKey identifies a model. A no is re-checked on the next call; if the underlying ChatModel already has a cached answer, that lookup is cheap.
Return
true if streaming is supported, false otherwise
Parameters
llm
The LLM operations instance
llm
Options used to resolve the LLM
Probes chatModel via StreamingCapabilityVerifier and caches a definitive answer.
A successful probe or UnsupportedOperationException is remembered. Other failures (missing key, network) answer false for this call but are not cached, and are logged once per instance so a provider outage does not present as a missing capability with no explanation.