Sacra estimates that Wrtn generated $33M in revenue in 2025, up 15x from $2.2M in 2024, as it introduced paid character chat in Korea and expanded its interactive storytelling model into Japan.

Revenue initially scaled through Crack, where paid character chat launched in October 2024 and monthly revenue increased from about $0.7M in November to $1.4M in December. Growth accelerated in 2025 with the launch of Kyarapu in Japan, with Crack and Kyarapu together generating more than $8M per month by April 2026.

The two apps reached that revenue on fewer than 500,000 combined monthly active users, implying more than $16 in monthly revenue per MAU. OOC, Wrtn's North American storytelling app launched in May 2026, reached $7.2M in monthly revenue in August, when overseas revenue surpassed domestic revenue for the first time.

Wrtn is targeting $145M in full-year 2026 revenue across Crack, Kyarapu, OOC and its early-stage enterprise AI business.

Wrtn closed its Series C on August 26, 2026, raising approximately $100 million. Valuation figures from that round conflict across sources, with reports citing approximately $722 million and others approximately $870 million. A reasonable midpoint is approximately $795 million post-money, based on the range of contemporaneous press accounts rather than a single confirmed figure.

Prior to the Series C, Wrtn raised a Series B of approximately $100 million in March 2025, led by Goodwater Capital. That round included Eugene Asset Management, Korea Development Bank, Capstone Partners, Woori Venture Partners, Sui Generis Partners, and Z Venture Capital.

Before the Series B, Wrtn completed a Pre-Series B in 2024 led by BRV Capital Management, with participation from Coreline Ventures and Antler Global. Earlier rounds included seed and pre-seed capital from Korean and international early-stage investors.

Across all rounds, Wrtn has raised approximately $166 million in total primary equity as of August 26, 2026.

Wrtn has two linked product layers. The first is a free, general-purpose AI app in Korea that acts as a consumer AI hub: users can run web-backed searches, upload documents or images for analysis, generate or rewrite text, create images, and use a personalized assistant that retains memory over time. The app routes queries across multiple frontier models, OpenAI, Anthropic, and Google, using a lightweight classifier to send simpler tasks to cheaper models and heavier tasks to more capable ones, giving users a single interface across providers.

The second layer is the revenue-generating product set. Crack in Korea, Kyarapu in Japan, and OOC in North America are interactive storytelling apps built on the same platform and localized by market. Users browse a ranked feed of story worlds and characters by genre, romance, fantasy, BL, sci-fi, simulation, pick one, and move the narrative forward through messages. The AI functions more like a dungeon master than a chatbot, generating text, scene images, and audio in response to user choices.

These apps differentiate from generic character chat through memory and mode design. The system tracks long-term relationship state, temporary context, user-defined world rules, and avatar identity across sessions, so a returning user resumes an existing narrative instead of restarting. Users can also adjust response quality and intensity by switching modes, with premium modes consuming more in-app credits and producing longer outputs.

Kyarapu's interface is structured like a content platform, with rankings, genre tags, creator titles, and visual novel sections, which addresses the blank-prompt problem common in AI chat products. OOC adds story and character creation tools, a certification program for top creators, and an Originals program through which Wrtn can acquire creator IP outright and publish it as first-party content.

The free Wrtn app and the paid storytelling apps function as a funnel. A user who starts with the free AI assistant for writing or search can discover Crack or OOC within the same ecosystem, then move into the higher-engagement, higher-spend entertainment layer without switching platforms.

Wrtn is a B2C consumer AI platform whose free general-purpose app serves as a top-of-funnel acquisition channel, while its storytelling apps, Crack, Kyarapu, and OOC, generate revenue. The free app is broad and low-friction, while the storytelling apps monetize through consumption-metered virtual currency.

In the storytelling apps, users receive a small free credit allowance and purchase additional credit packs when they run out. Every message, story action, and image generation consumes credits, and switching to a premium AI model tier or requesting longer outputs consumes more. Per-message credit prices have increased at least twice, in mid-2025 and mid-2026, indicating both higher demand and Wrtn's ability to raise pricing as the product matures. The model is closer to a game economy than a subscription: heavy users pay proportionally more, aligning revenue with the value extracted by the most engaged cohort.

This design has margin implications. Because the storytelling apps charge by consumption rather than offering unlimited access, Wrtn can route expensive inference to premium-tier payers while keeping lighter interactions cheap. The multi-model routing architecture reinforces this by reserving heavier frontier models for interactions users are already paying more to access. Wrtn also steers users toward web-based purchases over in-app purchases to reduce app-store commission leakage.

The creator ecosystem acts as a capital-efficient content supply mechanism. Rather than producing all story worlds in-house, Wrtn provides creation tools, discovery surfaces, and revenue-share rails that let users build the catalog. Wrtn selectively acquires high-performing creator IP through programs like OOC Originals, paying an upfront amount plus a share of credits generated, which gives it first-party content without the full cost structure of a traditional content studio.

The geographic expansion model reuses the same monetization playbook: prove the credit-consumption loop in Korea with Crack, replicate it in Japan with Kyarapu, then export it to North America with OOC. Each new market adds a monetization node without requiring a different product architecture. A nascent enterprise AI-transformation business exists alongside the consumer stack but remains early relative to the storytelling revenue engine.

Wrtn competes across two converging markets, general-purpose consumer AI and AI-native character entertainment. Competitive pressure differs by product layer, with entertainment rivals concentrated in character and fiction apps and utility rivals concentrated in global assistant platforms.

Character.AI is the closest benchmark for Wrtn's storytelling business in English-speaking markets. It has built a large creator ecosystem with millions of public characters, creator analytics, lore tools, memory upgrades, and, more recently, studio-produced micro-dramas, giving it a broader supply-side catalog than OOC currently offers.

Talkie, backed by MiniMax, reported roughly 20 million average monthly active users across its consumer AI products in the first nine months of 2025, making it a large distribution platform for AI entertainment rather than a niche app. PolyBuzz competes on breadth and permissiveness, with over 20 million characters and broad fandom indexing across anime, manga, and gaming IP, which pressures Wrtn's story products among users who prioritize variety over structured narrative systems.

Replika occupies a narrower but durable segment of the same behavior, persistent one-on-one emotional companionship with subscription-based monetization. Its focus on a single enduring relationship rather than a rotating cast of characters can produce stronger retention among users who want continuity over content variety, offering an alternative to Wrtn's entertainment-first framing.

ChatGPT, Gemini, and Perplexity are the most significant indirect threats to Wrtn's free utility app, which serves as its top-of-funnel acquisition layer. Gemini has crossed one billion monthly users and is adding voice, camera, and connected-service context. ChatGPT increasingly supports memory, project context, and companion-adjacent use cases by default.

The strategic risk is distribution rather than feature parity. If users satisfy their daily writing, search, and learning needs inside a global assistant with default phone placement, Wrtn must work harder to pull them into its own ecosystem rather than relying on the free app as an on-ramp to Crack or OOC.

NAVER is not a character-chat rival but is a material threat to Wrtn's consumer AI superapp ambition in Korea specifically. Its AI Briefing product reaches 30 million people monthly, and it is building action-oriented AI tied to its search, shopping, and creator graph, surfaces where it has proprietary Korean data and entrenched distribution that Wrtn cannot easily replicate.

CHAI represents a different structural risk: it is investing heavily in owned inference infrastructure, including a large GPU cluster and in-house models, which could allow it to compress costs and reinvest into free usage or richer premium tiers in ways that pressure Wrtn's margins in high-consumption storytelling.

Wrtn's expansion logic is geographic replication of a monetization playbook built on top of a free AI utility funnel that expands its addressable user base. The next vectors are new geographies, adjacent products, and higher-conversion content partnerships.

OOC's North American launch in May 2026 is the most immediate TAM expansion vector. Reaching $7.2 million in monthly revenue within three months of launch, with overseas revenue already exceeding domestic revenue by August 2026, North America indicates that Wrtn's credit-consumption model can travel outside Korea and Japan.

The anime and roleplay fandom communities that OOC targets in North America are large and already accustomed to spending on digital entertainment such as webtoons, visual novels, and gacha games, which reduces the behavioral friction of adopting a credit-based interactive fiction product. East Asian markets adjacent to Korea and Japan are a logical next step given Wrtn's localization infrastructure and its ability to adapt tone, genre mix, and community norms by market.

Wrtn Labs is developing Agent OS, a no-code platform for creating, deploying, and managing AI agents, with an agent store and connector store layered on top. If it gains traction, Agent OS would add a second revenue pillar in prosumer and SMB automation, a market structurally larger than consumer entertainment and one that can carry better margins.

The creator economy within Crack, Kyarapu, and OOC is also underdeveloped. As creator tooling matures and revenue-share mechanics become more transparent, Wrtn can attract professional writers, fan communities, and eventually licensed IP holders to build on its platform, shifting it from a content library toward a two-sided marketplace where Wrtn earns a take-rate on creator-generated spend.

A clear near-term conversion lever for OOC is pairing the platform with licensed anime, gaming, or webtoon IP. Wrtn already has the interaction engine and creator behavior in place. Official IP could increase conversion rates, reduce cold-start content risk, and make North American user acquisition more efficient by meeting fans inside existing fandom ecosystems.

Wrtn's multi-model partnership structure with OpenAI, Anthropic, and Google also creates a path to preferred access to new modalities, including voice, video, and richer image generation. That would expand what interactive storytelling means on the platform and open new credit-consumption surfaces beyond text-based narrative.

Model dependence: Wrtn's entire product stack, from the free utility app to the premium storytelling modes that drive nearly all revenue, runs on third-party frontier models from OpenAI, Anthropic, and Google, so any pricing increase, policy change, or access restriction at those providers flows directly into Wrtn's cost structure and product quality at a point when its storytelling apps are scaling rapidly and inference costs are already a primary margin constraint.

Fandom concentration: The April 2026 data showing more than $8 million in monthly revenue from fewer than 500,000 combined monthly active users across Crack and Kyarapu implies that a small, intensely engaged cohort of fandom users drives a disproportionate share of revenue, creating sensitivity to content trend shifts, platform moderation changes, or the arrival of a competing product that better serves that audience's tastes.

Safety and platform risk: As Wrtn scales OOC in North America, the AI companion and interactive fiction category is drawing more regulatory and app-store scrutiny around emotional attachment design, adult content boundaries, and minors protection, risks that could require heavier moderation that degrades immersion, constrains monetization mechanics, or triggers payment-rail restrictions as the North American market generates its fastest revenue growth.