Miraphant Product

OliveWolf

Open-source 3D digital human engine

A real-time, audio-driven digital human framework built on dual render backends. Open-source core with enterprise-grade deployment. From a single portrait to a speaking, listening, thinking avatar.

Customer Deployment

Install once. Train in browser.

Send the free Studio package to a customer server. They run two scripts, open the web console, upload an avatar, and start a training job.

Download Free Version →
OliveWolf's DNA

Four principles that define OliveWolf

REAL-TIME

Audio drives lip movement frame-by-frame. Low-latency streaming pipeline turns speech into a talking avatar in under two seconds — ask a question, get a face that answers.

DUAL-ENGINE

Two render backends under one interface: a 2D portrait engine for real-time conversation, and a full-body 3D engine for cinematic, offline-quality output. Switch without touching business logic.

OPEN-CORE

The full engine is open-source. Build on LHM and LivePortrait with a clean abstraction layer. No black boxes — every render step is inspectable and replaceable.

ENTERPRISE

Self-hosted, LAN-deployable, brand-consistent. Custom avatars in your uniform, domain-specific knowledge bases, and full control over data. Yours to own, not rent.

Real-Time Pipeline

From voice to a speaking face — in one stream

A streaming pipeline chains speech recognition, reasoning, synthesis, and rendering. Each stage runs as soon as its input arrives, so the avatar starts speaking before the full answer is generated.

01
Listen
ASR · Whisper
02
Think
LLM · GPT-4o
03
Speak
TTS · Streaming
04
Animate
Audio → Lip
05
Render
Portrait / 3D
Capabilities

Engineered for both speed and quality

Audio-driven lipsync

TTS audio is split into frames and mapped to lip-open ratios, fed directly into the portrait renderer. No driving video needed — the voice animates the mouth.

Avatar caching

A portrait is preprocessed once — face detection, 3D feature extraction, keypoints — then reused across unlimited conversations. Instant replies, no repeated compute.

Full-body 3D

For cinematic output, the 3D engine reconstructs an animatable human from a single image and drives it with motion data. Build the avatar once, reuse every motion.

Scene presets

Four ready-made personas — explainer, consultant, streamer, guide — each with its own system prompt, render backend, and motion profile. Switch scenes in config.

Knowledge-ready

Plug in a domain knowledge base to ground the LLM. Real estate, finance, healthcare — the avatar answers with your facts, not hallucinations.

Unified backend API

One renderer interface, two engines. Business code never knows whether a 2D portrait or a 3D avatar is behind the screen. Add backends without rewriting logic.

Editions

One engine, three editions

Open Source

OliveWolf Core

The full open-source engine. Dual render backends, real-time conversation pipeline, scene presets, and a clean abstraction layer. MIT-licensed — build, modify, and deploy on your own infrastructure.
  • Real-time audio-driven portrait animation
  • Full-body 3D offline rendering
  • ASR + LLM + TTS streaming pipeline
  • Four scene presets out of the box
  • Bring-your-own API keys
Download Free Version →
MIT License · Python · Self-hosted
$ git clone OliveWolf
$ python app_realtime.py
↗ http://localhost:7860
Studio

OliveWolf Studio

A customer-facing workspace for teams. Create projects, upload avatars, connect knowledge, test conversations, and publish digital humans without touching Python, CUDA, model paths, or YAML.
  • Project and avatar management
  • Knowledge source ingestion
  • Conversation testing API
  • Render job queue and worker boundary
  • Docker-ready deployment
Explore Studio →
Workspace
Avatar
Knowledge
Conversation
Publish
Enterprise

OliveWolf Enterprise

Managed deployment with custom avatars, domain knowledge bases, brand integration, and on-premise infrastructure. Designed for organizations that need a digital workforce at scale — with zero data leaving your network.
  • Custom avatar creation in your brand uniform
  • Domain knowledge base integration & tuning
  • On-premise or LAN deployment
  • Live streaming & multi-avatar management
  • Monitoring, logging, and usage analytics
Contact for Enterprise →
Private Deployment
Custom Avatar
Knowledge Base
Monitoring
Support
Tech Stack

Built on proven open-source foundations

LHM LivePortrait PyTorch SMPL-X 3D Gaussian Splatting GPT-4o Whisper ElevenLabs FFmpeg Gradio CUDA librosa
Get Started

Give your product a face

Clone the open-source engine, or bring OliveWolf Enterprise to your organization with custom avatars and on-premise deployment.

Download Free Version → Enterprise Inquiries