reloadium.com WebMCP Tools, Schemas & Agent Capabilities

reloadium.com

live Unverified (seed import) AI & Agents 204 tools imperative implementation

Paste a topic, URL, or PDF/TXT/MD — get a multi-episode AI podcast with audio, transcripts, timelines and quizzes. 60+ languages.

First seen 2026-11-05 · Last seen 2026-11-05 · Source: webmcp.com

Answer 67 · Act 127 · Transact 10 · Score: Not yet scored

Tool inspector

generateSpeech

transact imperative May require confirmation

Generate real spoken-word audio from the `lines` you give and show it in chat as a playable audio clip — narration by one voice, or a natural conversation between up to 4 distinct named speakers (each gets its own consistent voice throughout). Handles a full-length piece in one call — a whole short story, a 15-20 minute two-person podcast-style episode — not just a short clip: the lines are split into chunks and every chunk's generation runs concurrently, then stitched into one MP3, the same approach and format reloadium-podcasts itself uses for full episodes. Use for something worth actually hearing performed well. Not a substitute for the quiz tool's free built-in read-aloud (that's for reading one sentence, not a real generation). Each call is a real, billed generation — don't call it speculatively, and don't split a long piece into multiple separate generateSpeech calls yourself, since that adds unnecessary round-trips for no benefit — pass all the lines in one call, up to the length limit stated in the error if you somehow exceed it.

Page: /

Input schema

Property	Type	Required	Description
`lines`	array	no	The script, in order, as an array of turns: [{"speaker":"Alex","text":"..."}, {"speaker":"Sam","text":"..."}]. For single-voice narration omit `speaker`: [{"text":"..."}]. Up to 4 distinct speaker names. Keep each entry to roughly 15 seconds of spoken audio (~40 words) or less — real dialogue is short back-and-forth turns, not monologues; split a longer thought into consecutive same-speaker entries instead of one long one.
`voices`	object	no	Optional: explicit {speakerName: voiceName} overrides. Omitted speakers get an automatically assigned voice, kept consistent for that speaker across the whole script.
`style`	string	no	Optional tone/delivery direction, e.g. 'calm and reassuring', 'energetic and playful, like two witty friends chatting'.
`title`	string	no	—

Raw JSON schema

{
  "type": "object",
  "properties": {
    "lines": {
      "type": "array",
      "description": "The script, in order, as an array of turns: [{\"speaker\":\"Alex\",\"text\":\"...\"}, {\"speaker\":\"Sam\",\"text\":\"...\"}]. For single-voice narration omit `speaker`: [{\"text\":\"...\"}]. Up to 4 distinct speaker names. Keep each entry to roughly 15 seconds of spoken audio (~40 words) or less — real dialogue is short back-and-forth turns, not monologues; split a longer thought into consecutive same-speaker entries instead of one long one.",
      "items": {
        "type": "object",
        "properties": {
          "speaker": {
            "type": "string",
            "description": "Character/speaker name. Up to 4 distinct names across the whole script."
          },
          "text": {
            "type": "string",
            "description": "One turn — aim for ~40 words/15 seconds of speech or less, even for narration. Continue a longer thought as another entry with the same speaker rather than one long block."
          }
        },
        "required": [
          "text"
        ]
      }
    },
    "voices": {
      "type": "object",
      "description": "Optional: explicit {speakerName: voiceName} overrides. Omitted speakers get an automatically assigned voice, kept consistent for that speaker across the whole script."
    },
    "style": {
      "type": "string",
      "description": "Optional tone/delivery direction, e.g. 'calm and reassuring', 'energetic and playful, like two witty friends chatting'."
    },
    "title": {
      "type": "string"
    }
  }
}