v0.8.0

0.8.0 (2026-08-12)

Speak v2 Flux TTS streaming controls, Listen v2 redaction, a provider/version retype, and generator-level HTTP improvements. This is a breaking release — see the migration guide.

⚠ BREAKING CHANGES

• regen: AgentV1UpdateListenListen.provider is now AgentV1UpdateListenListenProvider (V1/V2 union); Google.version is now Optional<GoogleThinkProviderVersion>; SpeakV2SpeechMetadataControlsApplied has a new required breaksApplied field. See docs/Migrating-v0.7-to-v0.8.md.

Features

• Speak v2 (Flux TTS streaming): barge-in and reconfigure via sendInterrupt() / sendConfigure(), with onSpeechInterrupted / onConfigureSuccess / onConfigureFailure callbacks; new speed and expressivity connect parameters; new Deepgram Flux TTS voices. (#92) (ec519da)
• Listen v2: redact connect parameter (ListenV2Redact: numbers, aggressive_numbers). (#92) (ec519da)
• Core HTTP (generator 4.10.1 → 4.16.0): automatic response decompression by default (ResponseDecompressionInterceptor; gzip/deflate transparently decoded), and new optional retry tuning on ClientOptions.Builder (initialRetryDelayMillis, maxRetryDelayMillis, retryJitterFactor, all defaulted). (#92) (ec519da)

Compatibility

• A provider payload that omits the optional version discriminator (what 0.7.x emits) now deserializes as V2 (defaultImpl = V2Value) instead of dropping to {"provider":null}, so existing agent configurations keep working.