Expand description
The streamed output vocabulary: one CapabilityChunk per thing a runtime
emits — text, separated thinking, timestamped segments, audio, embedding
vectors, free-text status, and a terminal Done carrying generation stats.
Structs§
- Audio
Frame - A chunk of a raw binary audio payload plus the sample rate it was produced at (captured from the sidecar’s ready handshake).
- Generation
Stats - Metrics reported when a generation finishes. Every field is optional — a runtime fills in what it can measure.
Enums§
- Capability
Chunk - One item in a capability stream (chat, vision, speech, transcription, embeddings).