</>MCP Agents Market
Agent

CyberVerse

by Lynpoint1.6kPythonUpdated 2026-08-05

Self hosted, real-time digital human agent platform. Build voice-first AI agents with WebRTC, persona memory, tools, RAG, and optional digital-human video.

CyberVerse is an open-source, self-hosted platform for building real-time voice-first AI agents with optional digital-human video capabilities. It uses WebRTC for low-latency streaming, persona memory for conversation continuity across sessions, and a multi-agent architecture that delegates long-running tasks to background SubAgents while maintaining fluid voice interaction. Developers can create AI assistants ranging from pure voice agents to full video avatars that animate from a single photo, with pluggable modules for brain (LLM), voice (TTS/ASR), tools, memory, and face (avatar models). The platform supports both local GPU inference (FlashHead, LiveAct) and cloud digital-human APIs (Baidu Xiling, Xunfei, Vidu S1), with integrated RAG for knowledge-base retrieval and a web UI for runtime configuration.

Key Features

Real-time voice interaction via WebRTC with full-duplex support for natural interruptions and overlapping speech
Optional digital-human video from a single photo using local models (FlashHead 1.3B, LiveAct 18B) or cloud APIs (Baidu Xiling, Xunfei, Vidu S1)
Multi-agent architecture: PersonaAgent handles foreground conversation while SubAgents execute background tasks (search, research, summarization, HTML reports)
Persistent character memory and retrieval-augmented generation (RAG) for knowledge bases, documents, and biographical material
Plugin-based stack with swappable modules for LLM, TTS, ASR, embedding, and avatar; runtime config via web UI at /settings
LiteLLM integration for 100+ LLM providers (AWS Bedrock, Azure, Vertex AI, Mistral, Cohere) through unified interface
Voice-only mode (no GPU required) or GPU-accelerated avatar inference with real-time performance monitoring (RTP metrics)
Cloud image deployment on Compshare and AutoDL for rapid setup without manual dependency installation

Use Cases

  • 01Building a J.A.R.V.I.S.-style personal AI assistant that sees, hears, and responds in real-time video calls
  • 02Creating digital memorial avatars that preserve the voice and appearance of lost loved ones
  • 03Developing interactive character agents for entertainment, education, or customer service with persistent memory
  • 04Deploying enterprise voice assistants that delegate complex research and report generation to background agents
  • 05Prototyping multilingual conversational AI with character-specific knowledge bases and persona consistency
  • 06Running privacy-focused, self-hosted AI agents with local GPU inference instead of cloud-only services

Related Agents

View more

hermes-agent

The agent that grows with you

Agent250.4k

agency-agents

A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.

Agent155.5k

openinterpreter

A coding agent for open models like Kimi K3

Agent68k

cline

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

Agent66.3k

CyberVerse — FAQ

What is CyberVerse?+

CyberVerse is an open-source platform for building real-time voice-first AI agents with optional digital-human video capabilities. It combines WebRTC streaming, persona memory, retrieval-augmented generation (RAG), and pluggable modules for LLM, TTS, ASR, and avatar models.

How do I install CyberVerse?+

Clone the repository, create a Conda environment with Python 3.10+, copy infra/config to config, add API keys to config/env (Dashscope or Doubao), run 'make setup' to install dependencies, then start the inference server, Go API server, and frontend in three separate terminals. For voice-only mode, set inference.avatar.enabled to false in config/cyberverse.yaml.

What are the prerequisites for CyberVerse?+

Node 18+, Go 1.25, Conda, Python 3.10+, FFmpeg, libopus-dev, libopusfile-dev, libsoxr-dev, and pkg-config are required. For voice-only mode, no GPU is needed. For digital-human video, you need a CUDA 12.8+ GPU, PyTorch 2.8, and avatar model weights (FlashHead or LiveAct).

Which API keys does CyberVerse need?+

At minimum, you need API keys for an LLM provider: either DASHSCOPE_API_KEY for Alibaba Cloud Qwen models or DOUBAO_ACCESS_TOKEN and DOUBAO_APP_ID for Volcengine Doubao models. Optional cloud digital-human services (Baidu Xiling, Xunfei) require their own credentials.

Is CyberVerse free?+

CyberVerse is open-source under GPL v3.0 and free to use. However, runtime costs depend on the LLM, TTS, ASR, and cloud digital-human APIs you configure; local GPU inference avoids those costs but requires compatible hardware.

Which AI clients work with CyberVerse?+

CyberVerse is a standalone platform accessed via web browser (default at localhost:5173) and does not integrate as a plugin into Claude Desktop, Cursor, or similar AI clients. It serves its own frontend and exposes an API server on port 8080.

How do I install CyberVerse?+

Open the source repository on GitHub and follow its README. CyberVerse is a agent — MCP Agents Market links you directly to the official repo.

Is CyberVerse free?+

CyberVerse is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches

Audio AgentsKnowledge & Memory AgentsProductivity & Workflow AgentsBest MCP ServersBest SkillsBest PluginsTop rated listingsClaude Code toolsBrowse all categories