I built a text to speech app that will take any text and rea...

Karnage

npub1r0rs5q2gk0e3dk3nlc7gnu378ec6cnlenqp8a3cjhyzu6f8k5sgs4sq9ac

hex

80dcc67d9df7689504e5d62367405c6fb850891e5ba1f60d576a4abf003bcffc

nevent

nevent1qqsgphxx0kwlw6y4qnjavgm8gpwxlwzs3y09hg0kp4tk5j4lqqaullqprpmhxue69uhhyetvv9ujuem4d36kwatvw5hx6mm9qgsph3c2q9yt8uckmgelu0yf7glruudvfluesqn7cuftjpwdynm2gyg0drn8p

Kind-1 (TextNote)

2026-06-27T03:42:41Z

I built a text to speech app that will take any text and read it. Once read, the results are cached so you never have to re-engage the model again.

This is all fine.

But when I ask codex to generate audio from an article - it refuses.

Kinda strange that it will create a tool to do it, but refuses do it it when asked directly.

原始 JSON

{
  "kind": 1,
  "id": "80dcc67d9df7689504e5d62367405c6fb850891e5ba1f60d576a4abf003bcffc",
  "pubkey": "1bc70a0148b3f316da33fe3c89f23e3e71ac4ff998027ec712b905cd24f6a411",
  "created_at": 1782531761,
  "tags": [
    [
      "client",
      "Halo"
    ]
  ],
  "content": "I built a text to speech app that will take any text and read it. Once read, the results are cached so you never have to re-engage the model again. \n\nThis is all fine. \n\nBut when I ask codex to generate audio from an article - it refuses. \n\nKinda strange that it will create a tool to do it, but refuses do it it when asked directly.",
  "sig": "e76076c0071c575b280c6c1ad1e43eb60cac02663c0e9707bb8645c9e7324fd00f81a61ebb069058efb177f692ff9b01eb3b0c07c94a97827c40fafebc78adb8"
}