English two-voice technique test · Kore + Puck

Listen first, then inspect the transcript and the request that produced this file.

Listen to the demo

EnglishTwo voices in conversation0:16

As requestedgemini-3.8-flash-tts · Studio (prebuilt) · No style sent

Human-reviewedNot rated yet

Accent notes record one reviewer’s impression. They are not a panel verdict.

Exact transcript

NoraHey, Sam, have you had lunch yet?
SamNot yet. I was thinking of stopping by that café |oh really?| near the station.
NoraThat sounds good. I'm hungry too, but I have to finish a report first.
SamNo problem, I'll wait. |oh really?| Even after one o'clock is fine.

Generation recipe

This is the request that produced the file. Voice identity and delivery style are separate. Other settings: Default (not sent).

{
  "model": "gemini-3.8-flash-tts",
  "input": [
    {
      "type": "user_input",
      "content": [
        {
          "type": "text",
          "text": "Hey, Sam, have you had lunch yet?",
          "annotations": [
            {
              "type": "speech_metadata",
              "speaker": "Nora"
            }
          ]
        },
        {
          "type": "text",
          "text": "Not yet. I was thinking of stopping by that café |oh really?| near the station.",
          "annotations": [
            {
              "type": "speech_metadata",
              "speaker": "Sam"
            }
          ]
        },
        {
          "type": "text",
          "text": "That sounds good. I'm hungry too, but I have to finish a report first.",
          "annotations": [
            {
              "type": "speech_metadata",
              "speaker": "Nora"
            }
          ]
        },
        {
          "type": "text",
          "text": "No problem, I'll wait. |oh really?| Even after one o'clock is fine.",
          "annotations": [
            {
              "type": "speech_metadata",
              "speaker": "Sam"
            }
          ]
        }
      ]
    }
  ],
  "response_format": {
    "type": "audio"
  },
  "generation_config": {
    "speech_config": {
      "speakers": [
        {
          "speaker": "Nora",
          "voice": "Kore"
        },
        {
          "speaker": "Sam",
          "voice": "Puck"
        }
      ]
    }
  }
}