local VAD collects a spoken utterance,
Athena Whisper transcribes it,
OpenClaw's normal agent-consult path answers with its configured model and tools,
Athena Qwen3-TTS returns PCM audio to the Talk client.

Long replies are synthesized incrementally. The first short phrase starts playing as soon as it is ready while the next phrase is generated in parallel.

The provider intentionally uses half-duplex audio: microphone input is paused while a response is being transcribed, generated, synthesized, or played. This prevents speaker feedback from aborting TTS. Spoken interruption (barge-in) is therefore disabled; wait until playback finishes before speaking again.

No public speech provider is used. modelProvider names an existing OpenClaw model provider whose Athena base URL and API key are reused at runtime; no second key copy is required. If it is omitted, the plugin checks athena, llama-cpp, and openai in that order.

Recommended talk.realtime configuration:

{
  "provider": "athena-talk",
  "model": "athena-local",
  "speakerVoice": "alloy",
  "mode": "realtime",
  "transport": "gateway-relay",
  "brain": "agent-consult",
  "providers": {
    "athena-talk": {
      "modelProvider": "llama-cpp",
      "language": "de",
      "vadThreshold": 0.018,
      "silenceDurationMs": 750,
      "prefixPaddingMs": 300,
      "maxSpeechSeconds": 45
    }
  }
}

The plugin requires OpenClaw 2026.9.4 or newer. Build and validate it before installation:

npm install
npm run build
npm run check
openclaw plugins install . --force --accept-capabilities
openclaw plugins inspect athena-talk --runtime --json

Restart the gateway once if the installation does not trigger an automatic reload. The managed plugin copy is stored in OpenClaw's persistent data directory, so normal image updates do not remove it. Hermes remains unchanged; both clients reuse the same OpenAI-compatible Athena speech endpoints.