Voxta Setup wizard
First-run installer that gets your essential services configured in a few clicks.
The wizard is Voxta's friendly on-ramp. On first launch it walks you through picking the four essential service types so you don't have to learn the Manage Services UI before having a first conversation.
The four service categories
The wizard prompts you to pick at least one service for each of these:
| Category | What it does | Examples |
|---|---|---|
| Text Generation | Generates the character's text replies (the "thinking" part). | OpenAI, Anthropic, Voxta Cloud, llama.cpp, ExLlamaV2 |
| Text-to-Speech | Synthesizes the character's voice. | ElevenLabs, Coqui XTTS, Kokoro, UnrealSpeech, Cartesia |
| Speech-to-Text | Transcribes your microphone input. | Deepgram, Vosk, WhisperLive, Azure Speech |
| Action Inference & Summarization | Specialized LLM calls for action selection and chat summarization. | Often the same LLM as Text Generation, can be a separate cheaper model. |
Local vs cloud picks
The wizard mixes local-only options (free, but require a competent GPU) with cloud options (paid, no GPU needed). Hover any service in the picker to see its requirements.
Easiest path — pick Voxta Cloud for all four. One API key, everything works, monthly Patreon credits cover it. See Voxta Cloud → Getting Started.
Skipping the wizard
The wizard is optional. Click skip / close on first launch and configure services manually via Manage Services at your own pace.
You can also re-run the wizard later if you reinstall Voxta or want to swap stacks.