voxta 

SDK for building Voxta modules — the abstractions and registration helpers for creating custom text generation, speech, vision, and chat augmentation services as Voxta plugins.

Common utility helpers shared across Voxta server components and modules.

Shared data models and message contracts used by the Voxta HTTP and WebSocket clients to communicate with a Voxta server.

Core abstractions and interfaces shared across the Voxta platform — the module, service, and registration contracts used when building Voxta integrations.

Shared utilities for launching and managing external processes (such as Python executables) from Voxta modules.

Shared helpers for building Voxta text-to-speech (TTS) modules, including text normalization and audio handling utilities.

Shared utilities for downloading and working with Hugging Face models in Voxta modules.

Shared utilities for Voxta modules that run Python code, including environment setup and dependency management.

SignalR and HTTP client for connecting to a Voxta server to exchange chat messages and stream audio.

Hosting infrastructure for building Voxta providers — apps that extend a live Voxta session by registering actions, reacting to messages, updating context, and sending messages.

Reader and writer for the .voxpak container format — a single-file archive with content-addressed deduplication, independently compressed blocks for random access into multi-gigabyte assets, and optional... More information

Storage-side integration for Voxta packages — serves a resource's assets straight out of the .voxpak it was imported from, with the pack cache, scratch-file provider and DI registration that goes with it.

Shared utilities for building text generation service modules.

Builds the PG edition of a .voxpak: removes the explicit parts an author marked, so a content hub can offer a safe download of an adult package without the explicit content ever leaving the server.