Synthesize text into an AudioChunk using the voice provider.
Cache key is (sha256(text), voice) — equivalent determinism to
(text, voice) without pinning raw text in the cache payload. Effects pass
through effectFn AFTER a cache hit and are never part of the key, matching
the locked-decision invariant from the Python port.
Synthesize
textinto an AudioChunk using the voice provider.Cache key is
(sha256(text), voice)— equivalent determinism to(text, voice)without pinning raw text in the cache payload. Effects pass througheffectFnAFTER a cache hit and are never part of the key, matching the locked-decision invariant from the Python port.