NeuRelease

A fast neural parser for torrent and release names.

NeuRelease is a fast, high-performance, multilingual neural parser for torrent and release names. It reads a name, extracts its fields - title, season and episode, year, quality, codecs, languages, release group and more - and classifies what the name is: a film, a series, music, a game, software, a book, and whether it is anime. Every value comes with a confidence score and the span it was read from.

It combines a character-level CNN with a Transformer encoder, running as int8 inference with runtime-dispatched SIMD kernels. No ML runtime, no model download: the wheel carries the compiled library and the weights, about 5 MB, and Parser() needs no paths.

Pattern-based parsers recognise known markers and guess the rest by position, so anything ambiguous - a number that may be a year or an episode, a word that may be a language or part of the title - is settled the same way every time, right or wrong. NeuRelease decides from context. It is trained on hundreds of thousands of real, labelled release names from a large torrent index - mostly English, with German, Spanish, French, Italian, Russian, Chinese and Japanese names as well.

pip install neurelease

Use

from neurelease import Parser

parser = Parser()
r = parser.parse("Ted.Lasso.S03E03.4-5-1.1080p.ATVP.WEB-DL.DDP5.1.H.264-NTb")

r.title                                # 'Ted Lasso'
r.season, r.episode, r.episode_title   # 3, 3, '4-5-1'
r.streaming_service, r.release_group   # 'ATVP', ('NTb',)
r.source == "WEB-DL"                   # True: enums equal their labels
r.year                                 # None: the name does not say
r.title.confidence, r.episode_title.confidence  # 1.00, 0.80: every value knows how sure the model was
r.to_dict()                            # {'title': 'Ted Lasso', 'season': 3, 'episode': 3, ...}

names = ["Blade.Runner.2049.2017.2160p.UHD.BluRay.x265-TERMiNAL.mkv",
         "Stray_v1.5-Razor1911",
         "El Joven Sheldon - Temporada 6 [HDTV 720p][Cap.604][AC3 5.1 Castellano][www.pctnew.org]",
         "【高清剧集网 www.BTHDTV.com】邻家哥哥给我爱[第05-06集][简繁英字幕].Brother.Next.Door.2024.S01E05-06.1080p.WEB-DL.H264.AAC-BTHDTV",
         "葬送のフリーレン 第28話 「また会ったときに恥ずかしいからね」 (1080p).mkv"]
releases = parser.parse_batch(names)   # one result per name, in order

SHOW = ("title", "alternative_title", "season", "episode", "absolute_episode", "episode_title", "year", "content")
for r in releases:
    d = r.to_dict()
    print({k: d[k] for k in SHOW if d.get(k) is not None})
# {'title': 'Blade Runner 2049', 'year': 2017, 'content': 'movie'}
# {'title': 'Stray', 'content': 'game'}
# {'title': 'El Joven Sheldon', 'season': 6, 'episode': 4, 'content': 'series'}
# {'title': 'Brother Next Door', 'alternative_title': '邻家哥哥给我爱', 'season': 1, 'episode': [5, 6], 'year': 2024, 'content': 'series'}
# {'title': '葬送のフリーレン', 'absolute_episode': 28, 'episode_title': 'また会ったときに恥ずかしいからね', 'content': 'series', 'anime': True}

parse_batch runs many names at once on several threads and returns them in input order. Titles and evidence keep the original script - Latin, Han, Kana, Cyrillic. Every field a result can carry is documented in docs/RESULT.md.

Compared with GuessIt

Measured on the same machine on 2026-09-11 with the shipped model (version 3), on 3,344 video validation names:

About 3.4× faster on one thread in this measurement, and roughly ten times in batch.

GuessIt's regression corpus is its own test suite: fixture strings such as FooBar.307.PDTV-FlexGet and filesystem paths, written to exercise its rules, with every input assumed to be a video. NeuRelease parses a single release name as found in real traffic and classifies it before assuming anything, so on this corpus it scores 683 to 804 - and on real names the ranking reverses. The 22 limitation cases are GuessIt's own documented failures, not a representative sample.

Method, exact model identity, scoring snapshot and reproduction: docs/GUESSIT_COMPARISON.md.

Wheels

Python 3.10 or newer. The wheels are tagged py3-none-<platform>, one per platform rather than one per interpreter, because the library is loaded through ctypes and speaks the stable C ABI.

Anywhere else, build from source: the C++ library, the C ABI and the build instructions are in the repository.

License

MIT. PCRE2 is statically linked into the bundled library, and its notice travels in the wheel as THIRD_PARTY_NOTICES.md.