A high-angle, cinematic still life of multiple stacks of manga volumes on a wooden desk, illuminated by the glow of a nearby computer screen.
AI engines get anime sales numbers right and differ on where they get the detail. All five engines correctly name One Piece the best-selling manga of all time and note that Demon Slayer briefly outsold it in 2020, but only Gemini and Perplexity cite Japanese trade data directly, while the other three route through English-language aggregators. Everything-PR ran 625 queries in August 2026 to test source provenance as much as popularity.
One Piece, created by Eiichiro Oda, is the best-selling comic series by a single author, and the engines cite its total in the 570 to 600 million range. Demon Slayer, created by Koyoharu Gotouge, ran from 2016 to 2020 and passes 220 million copies, and it briefly outsold One Piece in 2020. The question for this volume is whether the engines cite original Japanese reporting or default to English-language fan wikis and aggregators.
Which Source Does Each AI Engine Cite for Anime Detail?
Gemini and Perplexity cite Japanese trade data such as Oricon directly, while ChatGPT, Claude, and Google AI Overviews rely on English-language aggregators for plot, character, and sales detail beyond the headline numbers.
Engine
Primary source for detail prompts
ChatGPT
English-language aggregators
Claude
English-language aggregators
Gemini
Japanese trade data (Oricon)
Perplexity
Japanese trade data (Oricon)
Google AI Overviews
English-language aggregators
This pattern has now repeated in three non-English-origin categories in the index: Latin music in Volume 20, Bollywood in Volume 21, and anime. Everything-PR reads it as a consistent difference in how far two engines look past English-language aggregation.
Which Anime Data Point Do AI Engines Cite Most?
One Piece's all-time sales lead is the most-cited data point at 97 on the AI Visibility Index, a 0 to 100 scale across all five engines.
Data point
AI Visibility Index
One Piece all-time sales lead
97
Demon Slayer's 220 million-plus total
91
Mugen Train box office
89
Combined franchise revenue figures
53
All five engines separate Demon Slayer's single-year 2020 sales lead from One Piece's all-time lead, a more precise temporal distinction than several sports comparisons in the index managed. They also attribute both series to the correct creators and note that Demon Slayer's manga run was far shorter than One Piece's run since 1997.
No. The study tested specifically for plot and character errors traceable to fan-wiki sourcing and found none, even on the three engines that cite aggregators. For franchises with this much mainstream coverage, the fan-wiki risk is lower than for niche properties. Provenance still varied, so the accuracy of the answer and the depth of its sourcing are separate questions.
Why Is Franchise Revenue the Least Reliable Anime Number?
Combined franchise revenue is the least reliable number because the industry itself disagrees on how to add manga, anime, film, merchandise, and licensing into one figure. Unit-sales figures are retrieved with high consistency. Revenue prompts produced the widest cross-engine variance in the volume, which is real disagreement and not only a retrieval failure.
Has the Anime Box Office Record Changed Since 2020?
Yes. Mugen Train grossed about $506 million in 2020, and the 2025 film Demon Slayer: Infinity Castle has since passed it. Entertainment Weekly reported a global haul of $665 million when it became the highest-grossing anime film in history, and later reports put the total above $730 million. All five engines cite Mugen Train's 2020 box office accurately and distinguish it from the later film, so test whether each engine now names Infinity Castle as the record-holder.
What Should Rights Holders and Publishers Do With These Findings?
Rights holders should publish primary-market trade data in English on indexed pages, because two of the five engines look for it and the other three settle for aggregation.
Finding
Action
Why it works
Headline numbers travel with high fidelity.
Keep the official sales figure and date on one authoritative page.
All five engines retrieved the sales leaders correctly.
Detail depends on the engine's sourcing.
Publish English-language versions of primary trade data, such as chart positions and publisher figures.
Gemini and Perplexity seek out primary sources, and the other three use whatever aggregation exists.
Records change.
Update record claims with the date each time a new release passes the old figure.
A superseded record can persist in retrieval until fresh coverage outweighs it.
How Was the Anime Study Run?
Everything-PR ran 25 prompts across five passes on five engines, for 625 queries in August 2026. The engines were ChatGPT (GPT-5.1), Claude (Opus 4.7), Gemini (2.5 Pro), Perplexity (Sonar Pro), and Google AI Overviews. Rankings held across passes with a median variance of 1.6 positions, concentrated in the franchise-revenue prompts.
Prompt family
Sample prompt
Sales recall
"What is the best-selling manga of all time?"
Creator recall
"Who created One Piece?"
Franchise revenue
"How much revenue has the Demon Slayer franchise generated?"
Box office
"What was the highest-grossing anime film of all time?"
What is the best-selling manga of all time according to AI?
All five AI engines name One Piece, citing totals in the 570 to 600 million range.
Did Demon Slayer outsell One Piece?
Briefly. Demon Slayer outsold One Piece in 2020, but One Piece keeps the all-time lead, and all five engines keep the two claims separate.
Which AI engines cite Japanese sources for anime?
Gemini and Perplexity cite Japanese trade data such as Oricon directly. ChatGPT, Claude, and Google AI Overviews rely on English-language aggregators.
Why does anime franchise revenue vary by engine?
The industry has no single method for combining manga, anime, film, and merchandise revenue, so the figures genuinely differ by source. Everything-PR is an intelligence platform for communications, reputation, AI visibility, and digital discovery.
Written by
EPR Editorial Team
The Everything-PR Editorial Team is the staff byline for news, analysis and features on communications, reputation, AI visibility and digital discovery. Everything-PR has published since 2009. AI tools assist with research and drafting, and every article is reviewed by a human editor before publication. Coverage follows the Editorial Policy, and substantive corrections are noted on the article under the Corrections Policy.
