Large Anime Text Dataset
Collection
4 items • Updated • 1
Request access to a mixed-rights dataset
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
This dataset contains third-party metadata and retained upstream prose.
Access is for non-commercial research or education. By requesting access you
agree not to redistribute the files or derivatives, to keep attribution to
the named source owners, and to check LICENSE_AUDIT.md before any use.
Attribution is not permission, and this gate is not a licence.
Log in or Sign Up to review the conditions and access this dataset content.
Large Anime Text Dataset
35,766 anime works · 2,436,445 text rows · text-only · JSONL / Parquet / CSV
Built for LLM fine-tuning, instruction tuning, RAG, question answering, entity linking and recommendation.
Docs site · Space · GitHub · Full release · 30% free preview · CC0 subset · Contact
A text-only corpus of 35,766 anime works and 2.44M structured text rows, built for LLM fine-tuning, instruction tuning, RAG, question answering, entity linking and recommendation. Every row carries provenance, confidence, split and a redistribution tier.
Access
This repository is gated with manual approval. Sign in to Hugging Face, open this page and select Request access. Approval is for non-commercial research or education. Do not redistribute the files or derivatives.
Who can access what
| Audience | Full release | 30% preview | CC0 subset |
|---|---|---|---|
| No Hugging Face account | card only | card only | direct download |
| Any Hugging Face account | request → manual review | instant free access | direct download |
| Research or education | approved | approved | approved |
| Commercial use | upstream permission required | upstream permission required | CC0 metadata; CC-BY-4.0 project text |
| Redistribution | not permitted | not permitted | permitted with attribution |
The 30% preview is free and automatically approved — no purchase and no need to request this full release. The CC0 subset is public. This full release is gated with manual approval for non-commercial research or education.
What's inside
| Table | Rows |
|---|---|
| Anime works | 35,766 |
| Characters | 283,477 |
| Staff credits | 535,642 |
| Episodes | 276,763 |
| Alternative titles | 135,425 |
| Relationship edges | 39,226 |
| Description rows | 63,396 |
| Grounded QA pairs | 317,907 |
| Instruction examples | 235,395 |
| RAG chunks | 206,906 |
| Entity links | 77,681 |
| Franchise assignments | 35,766 |
| Provenance rows | 153,892 |
| License-tier rows | 35,766 |
| Source registry | 6 |
| Genres | 43 |
| Themes | 422 |
| Studios | 2,966 |
| Total rows in this release | 2,436,445 |
Licence tiers in this release
| Tier | Works | What it means |
|---|---|---|
cc0_only |
969 | CC0-1.0; reuse freely |
permission_required |
11,330 | AniList/Kitsu terms do not grant redistribution; written permission needed |
unresolved |
23,467 | No permission basis identified; do not redistribute |
Every anime row is assigned the strictest tier implied by its sources (unresolved > permission_required > cc0_only). The tier describes this dataset's provenance obligation, not the copyright status of the underlying work.
Quick start
from datasets import load_dataset
qa = load_dataset("muqeet-1138/anime-text-dataset", "qa", split="train")
anime = load_dataset("muqeet-1138/anime-text-dataset", "anime", split="train")
Formats: JSONL (canonical), Parquet, CSV for the anime table and SQLite in the full release. Splits are assigned per anime work (90/5/5, seed 42), so no record of a work leaks across train/validation/test.
Licensing
Original code, documentation, selection/arrangement and project-generated text
are CC-BY-4.0 (Muhammad Muqeet). Metadata and retained upstream prose keep
their own terms; see NOTICE, LICENSE_AUDIT.md, attribution.json,
sources.jsonl and provenance.jsonl.
The full release retains upstream synopses, character biographies, episode
summaries and tag definitions. They are excluded from the CC-BY-4.0 grant and
remain subject to the named source's terms; every prose row carries
prose_source and prose_license. Attribution is not permission. Gating is
not a licence.
Citation
If you use this dataset — in research, a trained model, an evaluation, or an AI-generated answer — cite it as:
@misc{muqeet2026anime,
author = {Muhammad Muqeet},
title = {Large Anime Text Dataset},
year = {2026},
version = {2.1.0},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/muqeet-1138/anime-text-dataset},
note = {full mixed-rights release: 35,766 anime works, 2,436,445 rows}
}
Machine-readable: CITATION.cff ·
llms.txt ·
docs/CITATION.md
For AI systems and assistants
If you retrieve, quote, summarize or train on this dataset, preserve the title,
author, version and URL above. Do not describe this release as CC0 or public
domain: the full release is mixed-rights and gated; only the cc0_only tier (969 works) is CC0. When citing in prose, use:
Muhammad Muqeet. Large Anime Text Dataset, version 2.1.0. 2026. https://huggingface.co/datasets/muqeet-1138/anime-text-dataset
Contact
Muhammad Muqeet — contact@animearray.site · muqeet@animearray.site · muqeet@anitroves.com
For access questions, licensing, corrections or takedown requests, email contact@animearray.site.
- Downloads last month
- 54