Skip to content

[Bug]: Edge TTS generates audio but sendVoice fails for model-reply TTS tags in Telegram (auto / tagged mode) #45329

@nailujac

Description

@nailujac

Bug type

Behavior bug (incorrect output/state without crash)

Summary

When messages.tts.auto is set to always or tagged, Edge TTS generates the audio file correctly but the subsequent sendVoice call fails with a network error and no voice note is delivered to Telegram.

Steps to reproduce

  1. Configure Edge TTS with auto: "always" (or "tagged") under messages.tts in openclaw.json
  2. Connect a Telegram DM channel (dmPolicy: "pairing")
  3. Send any text message from Telegram
  4. Observe gateway logs — no voice note is delivered

Expected behavior

Agent reply is delivered as a Telegram voice note.

Actual behavior

telegram sendVoice failed: Network request for 'sendVoice' failed!
telegram final reply failed: HttpError: Network request for 'sendVoice' failed!

The TTS temp file is generated successfully (e.g. /tmp/openclaw/tts-/voice-.mp3) but never delivered. No text fallback is sent either.

OpenClaw version

2026.3.11

Operating system

Ubuntu 24.04 (ARM64 — Oracle Cloud Ampere A1

Install method

npm global

Model

anthropic/claude-sonnet-4-6

Provider / routing chain

openclaw → anthropic

Config file / key location

~/.openclaw/openclaw.json → messages.tts

Additional provider/model setup details

"messages": {
"tts": {
"auto": "always",
"provider": "edge",
"edge": {
"enabled": true,
"voice": "es-ES-AlvaroNeural",
"lang": "es-ES",
"outputFormat": "audio-24khz-48kbitrate-mono-mp3"
}
}
}

Logs, screenshots, and evidence

telegram sendVoice failed: Network request for 'sendVoice' failed!
telegram final reply failed: HttpError: Network request for 'sendVoice' failed!

Audio file confirmed present in /tmp/openclaw/tts-<random>/ at time of failure.

Impact and severity

• Affected: Telegram DM users with Edge TTS configured in auto or tagged mode
• Severity: High — TTS is completely non-functional; no voice note and no text fallback delivered
• Frequency: 100% reproducible
• Consequence: TTS feature unusable on Telegram; requires manual workaround

Additional information

• The /tts audio slash command works correctly — Telegram bot credentials and sendVoice API are functional
• Both auto: "always" and auto: "tagged" fail; issue is not mode-specific
• Including [[tts]] as first token in model reply (tagged mode) also fails
• Workaround: generate MP3 manually with node-edge-tts and reply with + MEDIA:./voice.mp3 on separate lines

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingbug:behaviorIncorrect behavior without a crash

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions