This project provides an AI-powered long-form novel generator, Long-Novel-GPT (LNGPT), designed for writers seeking to create extensive narratives with LLMs. It leverages a RAG-based agent to modify existing text and update plot outlines based on user input, aiming to streamline the novel-writing process.

How It Works

LNGPT functions as an agent that integrates with Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). Users can import existing novels, which the system then "deconstructs" to extract plot points and character relationships, generating a plot outline. The agent allows users to provide feedback or request changes, retrieving relevant text segments and plot summaries to modify the content while keeping the outline synchronized.

Quick Start & Requirements

• Install: Docker is the primary deployment method.

  docker pull maoxiaoyuz/long-novel-gpt:latest
  docker run -p 80:80 --env-file .env -d maoxiaoyuz/long-novel-gpt:latest
  

• Prerequisites: A .env file is required for API key configuration. For local LLM integration, ensure the LLM service exposes an OpenAI-compatible API.
• Setup: Requires Docker and an API key for LLM services. Configuration involves creating and populating a .env file.
• Docs: Demo Use Guide (local access after running)

Highlighted Details

• Supports importing existing novels for rewriting.
• Allows selection of different LLM models and real-time cost tracking.
• Features a new UI and chapter creation/editing capabilities.
• Aims for one-click generation of complete novels in future updates.

Maintenance & Community

The project is actively updated, with recent versions (2.2, 2.1) introducing significant features. Future plans include a more refined editing interface and support for additional models like Wenxin and Doubao. Community contributions are welcomed.

Licensing & Compatibility

The project is distributed via a Docker image (maoxiaoyuz/long-novel-gpt:latest). No specific license is mentioned in the README, which may require clarification for commercial use or closed-source integration.

Limitations & Caveats

The README states that the online demo has thread limitations, and full million-word novel generation requires local deployment with custom API keys and thread configuration. The project's ultimate goal of one-click novel generation is still in progress.