Skip to content

MNBVC

A very large Chinese corpus for AI training data

Visit official site

What MNBVC is

The project is described as a very large Chinese corpus built to give AI rich Chinese-language data, covering mainstream culture as well as niche culture and internet slang.

What you can do with it

  • Use large-scale Chinese corpus data
  • Support natural language processing and machine learning research
  • Support the development of Chinese-language AI technology

Application scenarios

  • Train a Chinese-language chatbot
  • Support Chinese text mining and sentiment analysis
  • Serve as a training base for Chinese natural language understanding models

Corpus contents

  • Plain-text data covers news, essays, novels, books and magazines
  • Papers, dialogue lines, posts and wiki pages are listed
  • Classical poems, lyrics, product descriptions, jokes and chat logs are listed
  • The supplementary note compares it with the 40T data used to train ChatGPT

Pros & cons

✓ What we like

  • Many text genres are gathered in one corpus
  • The material states it supports natural language processing and machine learning research

! What to watch out for

  • The listed material is plain-text Chinese data
  • The corpus centres on Chinese-language content

FAQ

What does MNBVC stand for?

The material expands it as Massive Never-ending BT Vast Chinese corpus.

What kinds of text are included?

News, essays, novels, books, magazines, papers, dialogue lines, posts, wiki pages, classical poems, lyrics, product descriptions, jokes and chat logs are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review