Notes

How many Japanese light novels have an English release on record? We counted.

Updated 2026-08-19

We're starting a site that publishes Japanese web novels in English, translated with their authors' permission. Before kicking it off, we wanted to know: how many Japanese series actually get released in other languages?

We couldn't find any compiled stats, so we counted them ourselves from public databases. We only managed it for English, and only for series recorded as light novels.

Here's what we found, where Japanese web novels stand right now, and what we're trying to build with this site. Let's look at the side with English releases first, then the side without.

Only 4.4% of Japanese series have an English release on record

RanobeDB is a volunteer-run database tracking Japanese light novel series and their translated releases. We used its search filters to count the series.

There are 22,575 series with a Japanese release on record. Of those, just 1,000 also have an English release. That's 4.4%. [Counted 2026-08-17]

We checked a second list, too. English Light Novels tracks 693 series released in English. Its inclusion policy and its definition of an "English release" both differ from RanobeDB's, so we can't add the two figures together or treat them as a range. We're leaving them side by side as two separate counts.

On its own, 4.4% doesn't tell us much. So which series end up in that 4.4%?

English releases cluster around the longest-running series

RanobeDB also tracks how many Japanese volumes each series has, so we broke the count down by series length.

(Counted 2026-08-17)

The share rises from 4.4% overall to 20.0% among series with at least five volumes, 37.7% at fifteen or more, and 41.9% at twenty or more. That last figure is nearly ten times the overall share. It slips back to 39.5% in the final row, which holds only 38 series. Within this dataset, longer-running series are more likely to have an English release on record.

But look at it the other way. Even at thirty volumes or more, 60% have no English release on record, and more than half of the series with twenty volumes or more don't have one either.

This table doesn't tell us why this happens

So English releases cluster around higher volume counts. But why? The table leaves out too much for us to say.

First, volume counts aren't sales. Some series run for many volumes without selling much; others become massive hits and wrap up early. We used volume counts because it's the only measure of a series' run visible from outside.

Second, we don't know the timeline. We can't see when an English licence was signed, which means we can't tell the difference between "it ran for a long time, so it got an English release" and "an English release was arranged early on."

Third, the table mixes decades of publishing history, many genres and many publishers together. Licensing decisions also turn on things it can't show: distribution deals, anime plans, how comparable titles sold in English.

And everything so far is only about series that made it into print. That still leaves the web novels that haven't.

For unpublished works, we looked at reader response instead

Volume counts don't work for unpublished web novels because, well, there are no volumes. Instead we have to look at reader response.

Shousetsuka ni Narou, usually just called Narou, hosts about 1.24 million works (excluding derivative fiction, as of August 17, 2026). That raw number isn't much use here. A story abandoned after three chapters and someone's writing exercise both count as "one work."

So we filtered by total points, the combined score of ratings and bookmarks from Japanese-language readers. Points don't measure how good a story is, and they say nothing about how English-language readers would respond. But Anime News Network reports that publishers typically start scouting works around the 10,000 to 30,000 point mark, and treat anything over 100,000 points as a megahit. [Source: ANN, 2026-01-12]

We actually gave up on this count once. We checked Narou's public API first, and nothing in the data tells you whether a work got a print deal; listings also cap out at 2,000 results. So we wrote up the conclusion that we couldn't count these works reliably enough, laid out the reasons, and planned to scrap the section.

Turns out we were wrong. The site's own search page displays the total number of works above any point threshold you set. We found the numbers we needed just after we'd given up on the section.

(Counted 2026-08-17, derivative fiction excluded)

There are 18,184 works sitting above the 10,000-point scouting line. 952 have crossed the 100,000-point megahit threshold.

The English catalogue on record is about the size of one platform's top tier

Put the two counts next to each other.

Light novel series RanobeDB records with an English release, across publishers and regardless of where the work began: 1,000.

Works above the 100,000-point megahit threshold on Narou alone: 952.

The two are close enough that we double-checked them.

They're different populations, and the only thing we're comparing is the size of the two pools. But in these database counts, the number of series with an English release on record is roughly the same as the number of Narou works in that platform's top tier.

We couldn't work out how many of those 18,184 Narou works are actually in print. Narou's public API has no field for it, so we couldn't determine print status in bulk.

We counted the gap, but not the reasons behind it

The count shows a gap, but it can't tell us what keeps it there. In our own work finding stories, writing to authors and keeping translations going, we keep hitting the same three hurdles.

First, for unpublished works, permission has to come from the author. There's no publisher rights desk, so we track down contact details ourselves, confirm we're talking to the author, and set out our terms. Plenty of authors are reachable through a platform's messages or through social media, but every conversation still happens one at a time.

Second, translation cost scales with word count. If a person is reading and revising the text, more words mean more time and more money. Web novels run for dozens of chapters. The job doesn't end when one book is done; a serial keeps going, and so does the translating. Whether it's one 20-volume series or four 5-volume ones, the workload follows the total word count.

Finally, nobody looks for a story they don't know exists. And a story nobody looks for doesn't show up as demand anyone can see.

We're building a place for authors to publish to the world themselves

Right now we reach out to authors directly, and we only translate once we have a green light. We don't require a particular domestic ranking or a print deal.

So far we're hosting 8 works and 133 chapters. Next to 1,000 or 18,184, that's a tiny start. Even so, there are plenty of good stories with no English release on record. We read a lot of them before choosing our first eight.

Of the three hurdles above, translation volume is the one that's getting easier to handle. A machine produces the first draft, and a person reads and revises the whole text. That makes keeping up with a serial feasible, and brings the workload within reach of one person.

For now, we're the ones writing to authors, and final responsibility for the English text sits with us. In time, we want authors making those calls themselves, publishing for readers anywhere on their own terms. Language shouldn't be the thing that keeps a good story from finding its readers, and that is something we can help change.

How we counted, and the limits

Method

• RanobeDB. We read the result counts on the series list: Release language set to Japanese for the denominator, Japanese and English together with and for the numerator, and minVolumes varied for the rows by volume.
• English Light Novels. We counted the entries in the alphabetical list, minus navigation and share items.
• Narou points. We varied min_globalpoint in the site's own search and read the number of works returned. Derivative fiction is excluded by default.
• All counts were taken on August 17, 2026.

Limits

• The counts move daily. The same filters a day later, on August 18, 2026, returned 22,591 series with a Japanese release and 1,001 with an English one. Landing on exactly 1,000 was a coincidence, not a cap on the display.
• "On record" is doing real work in this article. A series with no English release on record may still be readable in English: out of print, published under a different title, missing from the database, or circulating as an unofficial fan translation. And a release on record may be digital-only or cover only part of a series, so it doesn't mean you can read the whole thing today.
• Both lists are volunteer-run, not official industry data. The percentages describe what's inside RanobeDB, and its volume counts may not match actual publication exactly.
• We don't know which way gaps in coverage push the percentages. Works with an English release are probably easier for an English-language database to track, but so are long-running, well-known Japanese titles.
• The volume brackets are cumulative, so the rows aren't independent, and the 39.5% row holds only 38 series, too few to read as a ceiling.
• The Narou figures cover one site, and the three hurdles above aren't measurements. We didn't count Kakuyomu or the others, and the hurdles are what we ran into building this.

The numbers will keep moving, and we'll update this article at the same URL.