Open in interactive viewer → charts, metric popovers & call review

iQIYI's Sixth Rebrand: 'Long-Form Video Social Media' — and the AI Math Behind It

Q3 2026 brought a non-GAAP profit, a rescaled content model, and a $637 million receivable that not one analyst asked about.
IQ · Earnings Call · 2026-11-16

The moats keep multiplying

Six months ago, on the February call, iQIYI's CEO described two pillars — decentralization and all-in AI. Three months ago, on the August call, he had four moats: “these are the 4 moats iQIYI have built over the past decade” — Chang Yu, Investor Relations Director · 2026-08-18. On Monday's Q3 call, the architecture is two established moats plus three new ones, and the company has a new name for itself:

We are shifting from a centralized long-form video platform to a long-form video social media, a community co-built by content creators and audiences, while long-form videos such as dramas, films, variety shows, and animations remain at the core.

Tim Yu, Founder, Director · 2026-11-16
That framing is genuinely new for this company. In the curated keyword set it lands at the very top of the quarter — social media is the #1 theme, with content supply and long form video immediately behind it. That is not boilerplate drifting upward; it is the company's entire self-description being replaced in one quarter. The problem is that a rebrand is free, and this one arrives alongside a business that shrank. Total revenue was RMB 6.0 billion, down 4% sequentially. Membership services, the anchor, was RMB 4.0 billion, down 1% sequentially. Content distribution collapsed 41% sequentially to RMB 400.9 million — the old B2B licensing model quietly dying while the new story gets told. The offset: total cost fell 8% sequentially to RMB 3.5 billion, which delivered non-GAAP operating income of RMB 25 million, versus a RMB 30.3 million loss in Q2. Proft, yes — but a profit manufactured almost entirely by cutting the cost line faster than the revenue line.

The AI economics finally clear the bar they set for themselves

This is the part of the call that actually earns attention. In February, management's timeline for AI-native commercial film was cautious: “the AI-led commercial film probably will emerge within the next 2 to 3 years” — Chang Yu, Investor Relations Director · 2026-02-26. Nine months later they claim to have shipped eight of them in a single quarter, with The Ferry Man series generating over RMB 14 million in revenue-sharing income across three 60-minute-plus chapters, each recouping its full production cost within the first week of release. If that holds up under scrutiny, it is the fastest walk-back of a cautious AI timeline I have seen from a Chinese content platform. The supporting numbers are specific rather than aspirational: AIGC cut production cost 50–80% and production cycles by 80% on two animation titles; a children's AIGC title cut cost and timeline by over 70%. In micro-dramas, AI-generated content was nearly 90% of content volume and nearly 80% of views in September. This is not a slide deck — it is a cost curve. But there is a crack in the arithmetic, and it is worth naming. Asked about short-form drama supply, the Chief Content Officer said: “For the full year 2026, we expect to launch about 190 revenue-sharing short-form dramas, which is a substantial increase from more than 220 in 2025 last year” — Chang Yu, Investor Relations Director · 2026-11-16 190 is not a substantial increase on 220. Separately, AIGC short-form dramas went from 16 in Q2 to 21 in Q3 — a 31% step — yet 2027 supply is guided to "double year-over-year." Growth claims are being made about a base that management is simultaneously describing as shrinking. That is the kind of slip that tells you the narrative is being written faster than the spreadsheet.

What management stopped saying

The clearest signal in this quarter is not in the prepared remarks — it is in what vanished from them. NadouPro is gone. A year ago it was the flagship: a studio-grade AI production platform, 10,000+ active creators, a standalone monetization vehicle, an international version in development. In Q3 the platform appears in a single descriptive paragraph with no metrics, no creator count, no commercialization path — and in the keyword data NadouPro is one of the sharpest decliners of the quarter. The tool that was going to be the company's AI moat has been demoted to plumbing behind the social-media story. Thailand is gone. For four straight quarters, Thailand was the "proving ground" — the market whose operating playbook would be replicated across Southeast Asia, the second globally recognized content category after C-dramas. On Monday, overseas membership growth was reported at "nearly 30%," against the 40% reported in August, blamed on a lighter C-drama slate in Latin America. Thailand drops out of the keyword set entirely this quarter. The international growth engine is decelerating precisely as it stops being discussed. The hit-parade bragging is gone. Enlightent data rankings still pepper the script, but the emphasis has shifted deliberately. Management now says the quiet part out loud: “As the scale and revenue contribution of revenue-sharing content continue to rise, we expect our reliance on individual hit titles to decline” — Tim Yu, Founder, Director · 2026-11-16. That is a real strategic claim — and it is also the sound a company makes when it can no longer reliably manufacture hits. The economics are being swapped from fixed-cost ownership to performance payout: “Our revenue sharing model shifts a portion of iQIYI's content costs... from the fixed investment side to performance-based payout” — Chang Yu, Investor Relations Director · 2026-11-16. Fewer swings, smaller swings, and a content cost structure that improves structurally even if the top line does not.

The balance sheet nobody questioned

The Q&A ran three questions, down from five in August and four in May. Two of the three were about content strategy; one was about short-form drama competitive advantage. For a company with a sub-$1 billion market cap, thin coverage is expected — but the silence around the balance sheet was notable. In the space of two sentences, the CFO disclosed that operating cash flow was RMB 165.1 million, “making the fourth consecutive quarter of positive operating cash flow” — Tian Ying, Financial Officer · 2026-11-16, and then added: “we have a loan principal USD 636.6 million receivable from PAG” — Tian Ying, Financial Officer · 2026-11-16. That receivable is roughly 80% of iQIYI's entire market capitalization, sitting inside prepayments and other assets. It was not explained, not questioned, and not reconciled to the RMB 4 billion of cash and short-term investments also disclosed. Meanwhile the company repurchased approximately 64.2 million ADS for USD 76.4 million under the March 2026 program — an implied average near $1.19 per ADS, and a capital return larger than the quarter's operating cash generation. Buying back ~10% of your market cap while carrying a receivable worth most of the rest is an unusual allocation, and one the call did not defend.

The verdict

The AI cost story is real and measurable, and it is the first time in four quarters that iQIYI's AI claims have arrived with revenue attached rather than a roadmap. The pivots are working on the cost line, and the content mix is genuinely diversifying — live action is giving ground to AI-native formats, and revenue sharing model titles are scaling into the hundreds. But every quarter this company produces a new architecture, a new label, and a new set of moats, and each one arrives attached to a smaller revenue base. "Long-form video social media" may well be the right strategic read on where AI-generated supply takes the industry. It is also the sixth consecutive quarter in which iQIYI has told investors that the next era has already begun — while total revenue declined 4% sequentially and the only thing anyone on the call wanted to discuss was content strategy. Watch the content cost line, not the label.