Using ElevenLabs in LobeHub
ElevenLabs is an AI audio company. Its text-to-speech models generate expressive speech in 90+ languages from a large voice library. In LobeHub, ElevenLabs is a text-to-speech provider only: it has no chat models.
Step 1: Obtain your ElevenLabs API key
- Sign up or sign in at ElevenLabs.
- Open API Keys and create a key.
- Copy the key and store it securely.
Step 2: Configure ElevenLabs in LobeHub
- Open App Settings → AI Providers and select ElevenLabs (type
Elevenin the provider search box to find it). - Paste your key into API Key. Changes are saved automatically, and the key is encrypted with AES-GCM before it is stored.
- Turn on the switch next to the provider name.
- Optional: if your ElevenLabs workspace uses a data-residency endpoint, enter it in API Proxy URL (for example
https://api.eu.residency.elevenlabs.io). Leave it empty to usehttps://api.elevenlabs.io.
Step 3: Choose models
The Model List shows the ElevenLabs speech models under the TTS tab. Eleven v4, Eleven Multilingual v2, and Eleven Flash v2.5 are enabled by default; Eleven v3, Eleven Flash v2, and Eleven Turbo v2.5 can be turned on in the same list.
ElevenLabs has no chat model, so the provider page shows no Check button. Verify the key by generating speech as described below.
ElevenLabs bills each synthesized character. Review the ElevenLabs API pricing before generating large amounts of audio.
Generate speech
Speech synthesis runs through the LobeHub text-to-speech API, which the LobeHub CLI (lh) and the OpenAI-compatible POST /api/v1/audio/speech endpoint both use:
- Voice — when no voice is given, the default voice Roger is used, which also works with free-tier keys. Pass
--voice <voice id>to use another voice from the built-in list or from your own ElevenLabs voice library.
Apart from the default Roger, the built-in voices come from the ElevenLabs Voice Library, which free-tier API keys cannot use through the API. On the free plan, keep the default or pass a voice you own (for example one created with Voice Design or Instant Voice Cloning) with --voice <voice id>, or upgrade to a paid plan. Roger is one of the ElevenLabs Default voices, which only accounts created before March 2026 can use, until ElevenLabs retires them on 2026-12-31. When Roger is not available to your account, LobeHub picks a voice from your account instead (your own voices first, then premade ones); this needs the voices_read permission on the key. If no usable voice is found, pass one with --voice.
- Length — each request is limited to 4,096 characters, because the audio is returned in a single response. Split longer text into several requests.
- Format —
mp3(default) oropus. - Speed — Eleven Multilingual v2 and the Flash and Turbo models accept a speed from
0.7to1.2. Eleven v3 and v4 do not accept a speed setting.
Troubleshooting
"ElevenLabs API Key is incorrect or empty, please check your ElevenLabs API Key and try again." — No key was found, or ElevenLabs rejected the key. Enter a valid key in App Settings → AI Providers → ElevenLabs, or ask the administrator to set ELEVENLABS_API_KEY on the server. When ElevenLabs rejected the key, its own message follows in parentheses, for example ElevenLabs API error (401): Invalid API key.
"Free-tier ElevenLabs keys cannot use Voice Library voices…" — Your key is on the free plan and the voice comes from the Voice Library. See the voice note above.
Quota or rate-limit errors — Your ElevenLabs plan has run out of characters or reached its concurrency limit. Check usage in the ElevenLabs dashboard.
For self-hosted deployments, see ELEVENLABS_API_KEY in the model provider environment variables.