HanaVerse

Product Introduction

1. HanaVerse is an open-source, self-hosted web interface that integrates Ollama language models into an interactive experience featuring Hana, a 2D anime character who visually reacts to user inputs and delivers spoken responses via text-to-speech (TTS). It transforms raw LLM interactions into dynamic, multimodal conversations by combining visual animations, voice output, and customizable prompts.
2. The core value of HanaVerse lies in bridging the gap between static AI interactions and immersive user experiences, enabling developers and enthusiasts to deploy engaging AI companions with minimal infrastructure overhead. It prioritizes accessibility through self-hosting, extensibility via open-source code, and personalization using Ollama’s model ecosystem.

Main Features

1. Ollama Integration: HanaVerse connects directly to Ollama servers, supporting models like llama3:8b, codellama:7b, and mistral:latest, allowing users to switch between LLMs for specialized tasks such as coding assistance or creative writing. The UI provides real-time model status updates (e.g., “Initializing…” or “Ready”) and error handling for seamless integration.
2. Animated Character Interaction: Hana, the 2D anime persona, displays context-aware facial expressions and gestures synchronized with conversation flow, enhancing emotional engagement. The TTS system generates voice responses in natural language, with adjustable speech parameters for pitch, speed, and tone.
3. Customizable Prompts and Self-Hosting: Users define system prompts to tailor Hana’s personality or domain expertise, while the self-hosted architecture ensures data privacy and offline functionality. The platform is deployable via Docker or manual setup, with full control over server configurations and resource allocation.

Problems Solved

1. Impersonal AI Interactions: Traditional LLM interfaces lack visual and auditory feedback, making interactions feel transactional. HanaVerse solves this by embedding a reactive anime character that humanizes AI communication through emotive cues and voice synthesis.
2. Target User Group: Developers, AI hobbyists, and anime enthusiasts seeking to create interactive AI applications for entertainment, education, or customer support. Researchers studying human-AI interaction dynamics also benefit from its multimodal framework.
3. Typical Use Cases: Deploying Hana as a virtual assistant for coding help, integrating her into Discord servers for community engagement, or using her TTS capabilities for audiobook narration or language learning tools.

Unique Advantages

1. Multimodal Interaction Framework: Unlike Ollama’s default CLI or basic web UIs, HanaVerse merges LLMs with anime aesthetics and TTS, offering a cohesive experience unmatched by standalone tools like SillyTavern or Oobabooga.
2. Open-Source Flexibility: The MIT-licensed codebase allows modifications to Hana’s animations, voice engine, or integration with third-party APIs, fostering community-driven enhancements. Pre-configured Docker templates simplify deployment.
3. Cost-Effective Self-Hosting: By avoiding cloud dependencies, HanaVerse reduces operational costs and latency, making it ideal for low-budget projects or environments with strict data governance requirements.

Frequently Asked Questions (FAQ)

1. How do I self-host HanaVerse? Clone the GitHub repository, install dependencies using npm, configure the Ollama server URL in the environment variables, and deploy via docker-compose up or manual node server setup. Detailed instructions are available in the project’s README.
2. Which Ollama models are supported? All Ollama-compatible models work, including llama3:8b for general-purpose tasks, codellama:7b for programming assistance, and mistral:latest for multilingual support. Ensure your hardware meets the model’s VRAM requirements.
3. Can I customize Hana’s appearance or voice? Modify the character’s animations by editing the sprite sheets in the public/assets folder and adjust TTS parameters (e.g., voice gender, speed) in the config.json file. Community-created themes and voice packs are supported.
4. Does HanaVerse work offline? Yes, once deployed locally, HanaVerse operates offline if the Ollama server and required models are installed on the same machine. Internet access is only needed for initial setup or model downloads.
5. How can I contribute to the project? Fork the GitHub repository, submit pull requests for bug fixes or features, or share custom themes/animations via the community Discord channel. Issue templates are provided for reporting bugs or requesting enhancements.