hello! the record of my calls are in english but the ai y translating to spanish in my transcript. | ElevenLabs | Page 1

AI-generated Answer for xime_39173

Short answer: make sure your agent/transcriber is not set to Spanish or auto‑switching. Here’s how to force English-only transcripts, depending on what you’re using.

If you’re using ElevenLabs Conversational AI (Agents) for the calls

Keep the agent in English only: in the Agent settings, remove any Additional Languages so the agent stays in English for the whole call. If you enable additional languages, the widget can prompt or switch languages. (elevenlabs.io)
Disable language auto‑switching: if you added the Language detection system tool, remove it to prevent automatic switching based on detected speech. (elevenlabs.io)
Lock the language via overrides: if you start conversations programmatically (web/widget/SDK), set agent.language to "en" via conversation overrides. Ensure overrides are enabled in the agent’s Security tab. (elevenlabs.io)
Batch calling: if you’re using Batch Calls, make sure the CSV doesn’t set the language column to "es". You can also set language to "en" there (and enable overrides). (elevenlabs.io)
Embedded widget: if you use the web widget, set language="en" so the UI and agent language align. (elevenlabs.io)

If you’re generating transcripts with the Speech‑to‑Text API (Scribe)

Pass language_code="en" when creating the transcript to force English; otherwise the model auto‑detects and may choose another language. (elevenlabs.io)
Scribe returns words in the language of the audio (it does not auto‑translate), so forcing language_code avoids mis‑detection. (elevenlabs.io)

If you tell me whether you’re using Agents (call transcripts in the dashboard) or the Speech‑to‑Text API, I can give exact clicks/code for your setup.

#hello! the record of my calls are in english but the ai y translating to spanish in my transcript.