Member Junction
    Preparing search index...

    A usage/telemetry update emitted by a realtime client.

    Deltas preferred: each emission SHOULD carry the incremental token counts for the response/turn that just completed (not a session-cumulative running total), matching the server-side RealtimeUsage contract — consumers accumulate. A driver whose provider only reports cumulative totals must convert to deltas before emitting.

    Per-driver availability (capability, not laziness — providers without usage events simply never emit):

    • OpenAI — emits per-response deltas from the GA response.done frame's usage payload.
    • Gemini — emits per-turn deltas from LiveServerMessage.usageMetadata (promptTokenCount / responseTokenCount are per-response counts, not cumulative).
    • ElevenLabs — the Conversational AI socket exposes no usage events; never emits.
    • AssemblyAI — the streaming STT socket exposes no token-usage events; never emits.
    interface RealtimeClientUsage {
        InputTokens?: number;
        OutputTokens?: number;
        Raw?: unknown;
    }
    Index

    Properties

    InputTokens?: number

    Input tokens reported in this update (a delta for the completed response/turn).

    OutputTokens?: number

    Output tokens reported in this update (a delta for the completed response/turn).

    Raw?: unknown

    The raw provider usage payload, for hosts that want provider-specific detail.