Voxta Voxta Cloud
Hosted AI backend for Voxta — language, voice, and transcription models without a local GPU.
Voxta Cloud is Voxta's hosted AI backend. Plug your Voxta Server into it with an API key and you get language models, text-to-speech, and speech-to-text running on our infrastructure — no GPU required on your end.
It's a credit-metered service. Subscribe directly with Voxta, support us on Patreon, or buy one-time credit packs — all from the portal. See Plans & Billing for the options.
Voxta Cloud is in beta and improves continuously. Some models and pricing details may shift as we tune the service.
What's here
Getting Started
Generate an API key on the portal and connect your Voxta Server in five minutes.
Plans & Billing
Subscribe with Voxta or Patreon, buy credit packs, and manage or cancel.
Credits
How credits work, what burns through them fastest, and how to track usage.
Patreon
Subscriber tiers and what credits each one includes.
Security
How we secure your data, what we store, and our subprocessors.
Privacy
Our privacy policy and your data rights.
Terms
Terms of service, refunds, billing.
Quick facts
- Hosted on Microsoft Azure in US and Germany datacenters.
- LLMs run via OpenRouter (and optionally RunPod for some models).
- Voice synthesis via ElevenLabs, Cartesia, and UnrealSpeech.
- Transcription via Deepgram.
- We do not store prompts or responses — only metered usage (model, credits consumed).