Open WebUI voice without a GPU: hosted text to speech and Whisper
Open WebUI's OpenAI audio engine works with any service that speaks the OpenAI audio API. Point it at Glyphcall Voice and you get read-aloud and voice input on any server, with nothing to install and no OpenAI account.
1. Get a key
Get an access key on the pricing page: $1 once (1,000 credits) or $3 a month (3,000 credits, about 60 hours). After paying you get a page that shows the key (it starts with adt_). That key is the API key in every app below.
2. Text to speech
- Open Admin Panel → Settings → Audio.
- Set Text-to-Speech Engine to
OpenAI(not "OpenAI Realtime"). - Fill in the fields below and save.
| API Base URL | https://glyphcall.com/v1 |
|---|---|
| API Key | adt_… (your key) |
| TTS Model | tts-1 |
| TTS Voice | nova |
| OpenAI Params | {"response_format": "mp3"} |
Any OpenAI voice name works (alloy, nova, onyx, shimmer…), and so do the 40 English voice names on the voice list. Leave Response Splitting on Punctuation: each sentence is then one short call at $0.01. A single piece of text over 2,000 characters is refused, uncharged.
3. Speech to text (voice input)
- Same page: set Speech-to-Text Engine to
OpenAI. - Fill in the fields below and save.
| API Base URL | https://glyphcall.com/v1 |
|---|---|
| API Key | adt_… (your key) |
| STT Model | whisper-1 |
The model name is accepted and ignored: transcription runs on Whisper large-v3-turbo. Open WebUI converts recordings to MP3 before sending them, which suits us.
The same settings as environment variables
AUDIO_TTS_ENGINE=openai
AUDIO_TTS_OPENAI_API_BASE_URL=https://glyphcall.com/v1
AUDIO_TTS_OPENAI_API_KEY=adt_your_key
AUDIO_TTS_MODEL=tts-1
AUDIO_TTS_VOICE=nova
AUDIO_STT_ENGINE=openai
AUDIO_STT_OPENAI_API_BASE_URL=https://glyphcall.com/v1
AUDIO_STT_OPENAI_API_KEY=adt_your_key
AUDIO_STT_MODEL=whisper-1
Settings already saved in Open WebUI's database win over these variables: check Admin → Audio after a restart.
Test the key first
If something does not work in the app, check the key and the endpoint with one call from a terminal:
curl https://glyphcall.com/v1/audio/speech \
-H "Authorization: Bearer YOUR_KEY" -H "Content-Type: application/json" \
-d '{"model":"tts-1","input":"Hello from Glyphcall Voice.","voice":"nova"}' -o hello.mp3
What it costs
| Call | Price |
|---|---|
| Text to speech, up to 300 characters | $0.01 (10 credits) |
| Text to speech, 301 to 2,000 characters | $0.08 (80 credits) |
| Speech to text, a file up to 5 MB | $0.01 (10 credits) |
| Speech to text, a larger MP3 or WAV up to 60 minutes / 25 MB | $0.05 (50 credits) |
A failed call is never charged. No account, no subscription unless you choose the monthly plan.
Questions
Do I need an OpenAI account? No. The key comes from Glyphcall Voice and is paid per call from prepaid credits.
Which voices can I use? OpenAI's names (alloy, ash, coral, echo, fable, nova, onyx, sage, shimmer, verse) map to similar natural voices; the 40 English voice names on the voice list work as they are.
Why do I get an error for very long replies? One speech request takes up to 2,000 characters. With Response Splitting on Punctuation, Open WebUI sends one sentence at a time and never hits that limit.
Other guides: LibreChat · AnythingLLM · Home Assistant · Obsidian Whisper · Any other app · OpenAI tts-1 shutdown · OpenAI whisper-1 shutdown · Voice list