Sixty five of 170 title records carry 2026, four times the 2025 count, and most of that gap is the catalogue catching the present, not a production boom.
A production sector does not grow sevenfold in twelve months. Most of that step is the catalogue, not the industry.
Sixty five of the 170 title records on this site carry 2026 in their release year field. That is 46 percent of the catalogue, against 14 for 2025 and 17 for 2024. Not a majority, but four times the year before it, and the reason for that gap is not one thing.
The field read is release_year_text. A record counts toward a year if that year appears anywhere in it. Seven records name more than one year, so the column below sums past 170. Nothing is weighted by episode count or audience, because the catalogue holds neither. The rule counts a 2010 novel reissued in 2022 in both years, and a series running from December 2025 to July 2026 in both. Two of the 65 are that second kind.
| Year | Records naming it |
|---|---|
| 2026 | 65 |
| 2025 | 14 |
| 2024 | 17 |
| 2023 | 8 |
| 2022 | 10 |
| 2021 and earlier | 24 |
This has to be said before the rest. A catalogue built recently records the present as it happens and the past only when someone goes back for it. A title airing now is entered while it airs. A 2021 title is entered when a reader asks for it, or when it proves to be the source of something newer.
The size of that effect shows in the air dates. Reading aired_start, which 100 of the 170 records carry, and counting only premieres in the January to August window, the catalogue holds 11 for 2024, 7 for 2025 and 52 for 2026. A production sector does not grow sevenfold in twelve months. Most of that step is the catalogue, not the industry.
The two effects cannot be separated from inside this data. There is no external count of BL and GL premieres by year, and no field recording when a record was added here. Everything below describes the 2026 block. None of it claims what the world produced.
All 65 are TV Series (41) or Web Series (15). Not one is a novel, web novel, manga, manhwa, anime or musical. The catalogue holds 12 web novel, 7 novel, 5 manga and 4 anime records, and all of them sit outside 2026.
That is the coverage argument in a second form. Print and animation records enter mostly as source material for a drama already listed, so they arrive dated to publication, usually years back. A year recorded live holds only what is watched live.
The mix inside the screen records has moved. Web series are 15 of the 65, or 27 percent, against 17 percent catalogue wide, and 15 of the site’s 20 web series records are dated 2026. Those 15 lean away from Thailand: 5 South Korean, 4 Chinese, 2 Thai, one Taiwanese, one Filipino, and two co-productions, one Vietnam with Thailand and one across Hong Kong, Taiwan and China.
Counting every country a record names, so co-productions count in each, the 65 records produce 61 country mentions.
| Country | 2026 records | All records |
|---|---|---|
| Thailand | 30 | 72 |
| China | 7 | 15 |
| South Korea | 7 | 14 |
| Japan | 6 | 26 |
| Taiwan | 5 | 12 |
| Hong Kong | 3 | 3 |
| Singapore | 1 | 1 |
| Vietnam | 1 | 1 |
| Philippines | 1 | 1 |
Thailand’s share falls slightly inside 2026, from 58 percent of all records to 54. Japan’s falls harder, from 17 percent to 11, mostly the print and anime records dropping out. Taiwan holds 5 of its 12 records here. The sharper line is lower down. Hong Kong, Singapore, Vietnam and the Philippines appear in this catalogue only in 2026, across six records: Ash, Kiss or Kill, Sammy’s Children’s Day, Double Helix, Sending the Wounds and Boyfriend Express. Four of the catalogue’s five multi-country records are dated 2026 as well.
Reading aired_end against 15 September 2026, the 65 split four ways. Thirty seven have an end date already past. Seventeen are mid run, start behind, end ahead. Backdoor has a start date and no end date recorded. The Crow Club has not premiered and is scheduled to run to 1 January 2027.
So 19 of 65, just over a third, are unfinished. The same test on every non 2026 screen record returns 29 of 29 finished. That is the cleanest structural difference here, and it is one of timing, not of kind. Two mid run entries stretch far: LoveX3 to 22 November and Mr. Fanboy to 14 November. Gelboys runs from 8 February 2025 to 26 September 2026, which is why it sits in both year columns.
The 65 records name 65 distinct production companies. The whole catalogue names 131, and 55 of the 65 appear in 2026 and nowhere else. Most of the catalogue’s studio list was written by a single year.
| Studio | 2026 | All years |
|---|---|---|
| GMMTV | 8 | 14 |
| Studio Wabi Sabi | 5 | 9 |
| Domundi TV | 4 | 4 |
| Mandee Work | 4 | 4 |
| The One Enterprise | 2 | 3 |
| Change2561 | 1 | 3 |
| avex pictures | 3 | 0 |
GMMTV holds 8 of the 65, the largest single share, on records including Moonshadow, Weirdo 101 and When Oranges Fall. Domundi TV and Mandee Work appear four times each and only in 2026, often together, as on Love Upon a Time and Your Third. Moving the other way, the 3 avex pictures records are all anime and all earlier.
The episode_duration field parses on 51 of the 65 and on 27 of the 29 non 2026 screen records. Median minutes per episode: 45 for 2026, 49 for the rest. Median episode count: 10 against 12.
The obvious explanation is the extra web series, short by design. It is incomplete. Restricting to TV Series only, the 2026 median is 47 minutes against 52 for earlier TV Series records, so the shift survives holding format constant on a small sample.
The spread stays wide. The shortest 2026 record, Fix the Error, lists 9 minutes an episode; Unexpectedly Naughty Fukami lists 13; Connecting to You lists 26; Love Upon a Time lists 69.
Sample: 170 title records, read on 15 September 2026 from the site’s own structured fields. Year from release_year_text, country from country_text, type from content_type_text, studio from production_company_text, air status from aired_start and aired_end, length from episode_duration and episodes. Multi value fields were split on commas, each value counted once per record.
One framing did not survive checking. It is tempting to say most of this catalogue is 2026, and it is wrong. Sixty five of 170 is 46 percent, close to half and short of it, so 2026 is the largest year here rather than a majority of it.
Three limitations. There is no field recording when a record was added, so the coverage effect can be shown but not measured. Twenty one records carry no aired_start, all of them print or animation, so the premiere comparison is a screen record comparison only. Five 2026 records carry no parsed episode duration and four of those are still airing, so the runtime medians will move.
One check returned nothing usable. The 2026 records show no January or February premieres, the earliest being 13 March. That looks like a scheduling finding and almost certainly is not one, since a real calendar would carry some. It is reported here as a gap in the data rather than a pattern.
The summary is narrow. The 2026 block is the largest single year here, entirely screen records, a third unfinished, naming 65 studios of which 55 appear nowhere else, reaching four countries no other year reaches, with episodes about four minutes shorter at the median. Each of those is a fact about the catalogue first and about production second.
This page will read differently in September 2027. The 19 unfinished records will have end dates. The 2026 count will rise as late year premieres are entered, then stop once the year closes. The 2025 and 2024 counts should rise too, slowly, as older titles are added behind the new ones. If 2027 lands at 50 while 2026 settles at 70, the gap starts to mean something about production. Until one settled year can be set against another, this number mostly measures how recently the instrument was built.
Sixty five of the 170 title records name 2026 in their release year field, which is 46 percent of the catalogue. That is the largest single year here, against 14 records for 2025 and 17 for 2024.
It cannot be read that way from this data alone. A catalogue built recently enters current titles as they air and older titles only when someone goes back for them, so the current year is always over-counted.
Every record’s release_year_text field was read on 15 September 2026, and a record counts toward a year if that year appears anywhere in the field. Seven records name more than one year, so the year column sums past 170.
It makes no claim about how many BL or GL productions existed in any year. There is no external premiere count and no field recording when a record was added, so the production effect and the coverage effect cannot be separated.
Counting every country a record names, the 65 records give Thailand 30, China 7, South Korea 7, Japan 6, Taiwan 5, Hong Kong 3, and one each for Singapore, Vietnam and the Philippines.
Nineteen of the 65 are unfinished as of 15 September 2026. Seventeen are mid run, one has a start date and no recorded end date, and one has not premiered yet.
GMMTV with 8, Studio Wabi Sabi with 5, and Domundi TV and Mandee Work with 4 each. The 65 records name 65 distinct production companies, and 55 of those appear in no other year.
Slightly. The median 2026 record lists 45 minutes an episode against 49 for earlier screen records, and the gap holds within TV Series alone at 47 against 52.
Metadata and images in this article come from TMDB and official sources, and remain the property of their copyright holders. This product uses the TMDB API but is not endorsed or certified by TMDB. BLGLFLIX is an independent site with no affiliation to the productions covered. Full copyright, correction and takedown policy: Disclaimer.
Discussion 0 comments
Join other fans talking about Why So Many Records Here Are Dated 2026. Be kind, no spoilers without tags, no spam.
No comments yet.
Share the first reaction below.