AI has given Volcano Engine a chance to change the table, but winning still follows the old rules.

Editor | Meng Wen

"In the past couple of years, everyone competed on who could sell Tokens cheaper. Now, it's about who can go deeper."

Generative AI has overturned the table that traditional cloud providers had been guarding for a decade, allowing Volcano Engine, which entered the market six years later, to rely on C-end traffic and ultra-low prices to become the leader in public cloud large model invocation volume.

However, as the industry's narrative shifts from 'Token worship' to 'production-grade implementation,' Volcano Engine suddenly finds that the heavy barriers related to assets, services, and trust—which were once bypassed by traffic—can no longer be avoided.

After capturing the dividends of the AI era, Volcano Engine has no choice but to enter a longer and more cumbersome business.

AI Reshuffles the Deck: Volcano Engine 'Joins the Table'

When Volcano Engine officially launched commercially in 2020, its goal was to become the 'fourth cloud' alongside Alibaba Cloud, Tencent Cloud, and Huawei Cloud. However, few in the industry believed it could succeed at the time.

It wasn't that ByteDance lacked technical capabilities, but rather that the moat of the traditional IaaS market had been built over decades.

Migrating a medium-sized enterprise's core business to the cloud takes an average of 6 to 18 months, with sunk costs ranging from millions to tens of millions of yuan. A survey of over 300 global enterprise IT leaders revealed that 57% of companies spent more than $1 million (approximately 7 million RMB) on migration, with an average project cost of $1.75 million (approximately 12 million RMB).

Under such strong stickiness, the market had already formed an unbreakable Matthew effect: IDC data from the second half of 2022 showed that Alibaba Cloud, Huawei Cloud, China Telecom Cloud, and Tencent Cloud together held over 70% of the public cloud market share, with Volcano Engine and other small and medium-sized providers lumped into the 'others' category.

ByteDance's be good at (expertise in) C-end traffic strategies proved ineffective here, as the hundreds of millions of users on Douyin and Toutiao could not assist B-end clients in cloud migration.

The turning point came with the generative AI boom triggered by ChatGPT at the end of 2022, which created a brand-new market segment—GenAI IaaS—with no historical market share precipitate (accumulation).

The core of traditional IaaS is CPU general-purpose computing power, competing on data center scale, network coverage, and cost control of general-purpose servers. However, the core of GenAI IaaS is GPU intelligent computing clusters, competing on the scheduling capability of ten-thousand-card-scale GPUs, the transmission efficiency of RDMA networks, and the stability of large model training—capabilities that no cloud provider had accumulated in advance.

AI large models essentially dealt a new hand to all players, allowing everyone to restart from the same line.

While other providers were still digesting their inventory of traditional general-purpose computing power and hesitating to heavily invest in GPUs, Volcano Engine went against the market trend and made large-scale purchases of computing resources, directly securing training orders from the first wave of large model startups such as Zhipu AI, Yuezhia'anmian, Minimax, and 01.AI. This propelled it to become the cloud provider with the highest share in the GenAI IaaS sector.

If breaking through in intelligent computing was about seizing the timing of the computing power structure shift, leading in MaaS was about Volcano Engine truly cashing in its C-end DNA as B-end leverage for the first time.

The large model industry has a natural flywheel effect: more users → more dialogue data generated → better model iteration → attracting even more users. Traditional cloud providers' MaaS businesses rely on gradual accumulation of invocation volume from B-end clients, making the flywheel turn extremely slowly.

However, Volcano Engine is backed by Doubao, a national-level C-end app with over 100 million daily active users, effectively acting as a natural traffic accelerator for its MaaS business.

In May 2024, Volcano Engine initiated a price war, reducing the enterprise pricing of Doubao's main models to 99.3% below the industry average, directly ushering in the 'cent era' for large model APIs. Combined with its C-end user base, the invocation volume of Doubao's large model Tokens soared.

By the FORCE conference, its daily average Token invocation volume had surpassed 180 trillion, growing over 1,500 times compared to two years ago. Calculated by public cloud large model invocation volume share, Volcano Engine ranked first in China with 49.5%.

Many assume Volcano Engine's top invocation volume is inflated by Doubao's C-end traffic. In reality, as an internal ByteDance business, the Token consumption generated by Doubao's C-end does not count toward Volcano Engine's publicly reported enterprise service invocation statistics. Volcano Engine's near-half share of the public cloud MaaS market relies on model capabilities and extreme cost advantages, not C-end traffic diversion.

Moreover, in the AI large model era, a strong model itself serves as a new 'entry point' for the cloud. ByteDance's self-developed Doubao series models—including the Seedance series, which unlocks commercial video production scenarios, and Doubao 2.1 Pro, which achieves production-grade Coding and Agent capabilities—are the strongest customer acquisition tools: clients seeking top-tier multimodal generation capabilities or code models that can directly integrate into R&D workflows will prioritize Volcano Engine. Meanwhile, clients' real-world scenario data feeds back into model iteration, forming a positive flywheel of 'strong model → customer inflow → scenario refinement → stronger model → more customer inflow.'

This aligns perfectly with Tan Dai's conference assertion that 'models can stably enter enterprise production workflows after crossing the production quality threshold': when model capabilities are strong enough to reliably handle core enterprise production processes, customer switching costs rise rapidly, creating user stickiness that rivals the barriers built by traditional cloud migration costs.

However, it must be acknowledged that this flywheel currently primarily covers internet clients, developer communities, and public cloud API scenarios. By relying on 'Token sales + strong models,' Volcano Engine has indeed achieved a overtake on a curve (overtaking on a curve) in the MaaS track (track) and driven overall public cloud market share gains.

Yet the 'first place' title has been controversial from the start due to differing statistical caliber (methodologies) and customer structures.

Calculated by public cloud API invocation volume, Volcano Engine is undoubtedly first. However, when measured by overall AI cloud revenue, Omdia data shows Alibaba Cloud remains first with a 35.8% share, its revenue exceeding the combined total of second to fourth place.

A more fundamental difference lies in customer structure. Currently, the domestic MaaS market still derives most revenue from public cloud API invocations, but high-value production scenarios such as financial risk control, autonomous driving simulation, government data governance, and industrial large model deployment predominantly adopt private deployment or dedicated cloud models, where single-customer ARPU values are tens or even hundreds of times higher than public cloud APIs. This market has long been dominated by Alibaba Cloud, Huawei Cloud, and Tencent Cloud.

AI has given Volcano Engine a chance to overtake by changing lanes, pulling it from the 'others' category in the traditional cloud market to the center of the table.

From Token Competition to Delivery Excellence

The cloud market's wind direction (wind direction) has shifted this year.

If 2023-2024 were defined by 'land grabbing'—competing on model quantity, price, and invocation volume—the keyword for the second half of 2025 has become 'production implementation.'

Enterprise clients no longer pay for 'cheap Tokens' but instead ask tough questions: Can large models solve my real problems? Will my data leak? Can they integrate with my existing systems? Who takes responsibility if something goes wrong?

At this industry turning point, Volcano Engine held the FORCE conference, signaling a shift: it would no longer compete on 'who is cheaper' but instead proactively address the capabilities needed for core production scenarios.

The reassurance came at the organizational strategy level. At the conference, ByteDance CEO Liang Rubo stated: Scaling AI peaks is ByteDance's top priority, and Volcano Engine's MaaS business has been upgraded to a group-level foundational business with long-term, firm (firm) investment.

In the cloud industry, upgrading a business from 'line-of-business level' to 'group foundational business' is the highest-weight strategic signal, meaning MaaS is no longer just a Volcano Engine department matter but ByteDance's infrastructure: group-level computing power procurement, talent allocation, and capital investment will fully tilt . Even internal AI demands from Douyin, Toutiao, and Feishu will prioritize Volcano Engine for practice.

When Huawei Cloud upgraded its EI Enterprise Intelligence to a group core business in 2017, it captured the government cloud market first within three years. When Alibaba Cloud upgraded to the Alibaba Cloud Intelligence Business Group in November 2018, led by Group CTO Zhang Jianfeng, it maintained its long-term first-place share in China's public cloud market. Volcano Engine's adjustment similarly greenlights long-term heavy investment.

Next came addressing the weak link of 'enterprise-level capabilities.' Following Volcano Engine President Tan Dai's judgment that 'models must cross the production quality threshold to handle enterprise business,' Volcano restructured its three-tier AI cloud-native service architecture: the MaaS layer launched Ark CLI to lower access barriers, the Agent layer upgraded AgentKit and HiAgent 3.0, and introduced the enterprise workstation ArkClaw.

The most critical addition was the AI Trust security system at the security layer, addressing government and enterprise data compliance pain points through confidential computing, intelligent security operations, and full-chain auditing. It even jointly built an AI confidential computing zone with China Mobile, directly targeting central SOEs' core requirement of 'data not leaving the domain.'

After the introduction of the Data Security Law and Critical Information Infrastructure Security Protection Regulations, 'data not leaving the domain, full-chain auditable' has become an ironclad entry threshold in government, finance, and energy sectors. Previously, Volcano Engine lacked systematic trusted computing capabilities, making it difficult to even qualify as a supplier for many central SOEs, let alone secure core orders.

Partnering with China Mobile was the right move: operators are the most trusted entry point in government and central SOE cloud markets. Huawei Cloud and Alibaba Cloud expanded into government and enterprise (government and enterprise) markets early on by binding with operators. Volcano Engine's proactive 'operator binding' paves the way for entering high-value government and enterprise markets.

Beyond enterprise capabilities, Volcano Engine also began addressing doubts about its business model. At the conference, Doubao fully launched a tiered subscription system: Basic at 68 RMB/month, Enhanced at 200 RMB/month, and Professional at 500 RMB/month. Core functions like daily chat and basic Q&A remain permanently free, with fees only for advanced features like long document processing, AI office tools, and high-definition generation.

Previously, doubts about Doubao's monetization persisted. LatePost reported that Doubao's daily revenue was once below 1 million RMB, mostly from e-commerce referral commissions, while its B-end API business had initiated a price war, leading to persistent high computing costs amid skyrocketing Token invocations—voices of 'more traffic, bigger losses' lingered.

This paid tiering means more than just earning C-end membership fees. Morgan Stanley's estimates for domestic large model monetization show that long-term continuous subscription rates for C-end large models range from 0.3%-3%, with a neutral expectation of ~1% for top products. Tencent Research also found that including single-function payments and temporary top-ups, domestic AI users' overall payment rate is ~9.8%.

Based on QuestMobile's reported 345 million monthly active Doubao users, even achieving a stable 1% subscription rate could generate hundreds of millions in annual stable cash flow. This revenue can directly support B-end R&D and price wars, reducing reliance on Tier 1 market financing or traditional business subsidies—unlike other large model providers.

This marks the first time a leading domestic large model provider has directly connected C-end payment systems with B-end MaaS businesses. Once the payment model proves viable, Volcano Engine will become one of the few players in the industry capable of using stable C-end cash flow to support B-end R&D and price wars, essentially submitting a pledge for MaaS business sustainability.

Meanwhile, the criteria for measuring business success at Volcano have also started to change. At this conference, the organizers did not repeatedly emphasize being 'number one in daily average Token calls.' Instead, they highlighted 'over 200 enterprise clients with annual cumulative Token consumption exceeding RMB 1 trillion, doubling in half a year' as a core metric. They specifically mentioned that 90% of the world's leading mobile phone manufacturers, 100% of mainstream automotive companies, and over 80% of systemically important banks have already integrated Volcano Ark.

According to the general implementation framework proposed by the Ministry of Industry and Information Technology's 'Guidelines for Promoting Enterprise Cloud Adoption,' enterprise cloud usage typically goes through three stages: 'single-point testing - partial deployment - core system migration.' When AI capabilities are only used for scattered efficiency-enhancing scenarios like chatting and content generation, customers face almost no migration costs. However, when AI penetrates deep into a company's core production processes, system coupling increases significantly, and customer retention rates also rise notably.

This customer retention metric carries more weight than mere call volume numbers, indicating that Volcano has at least moved beyond the 'customer trial' phase.

All these changes ultimately reflect in growth targets. Volcano Engine's internal revenue target for its MaaS business in 2026 has been raised to the RMB 10 billion level, representing severalfold growth from 2025. Tan Dai confirmed in an interview that the revenue target has indeed been raised this year, with a consistent pursuit of 'profitable scale.'

Volcano Engine is no longer satisfied with being a platform leading in Token call volume. Instead, it aims to become an AI cloud company capable of sustained profitability and stable delivery.

Volcano Engine Embraces the 'Tough Business'

Looking back at Volcano Engine's growth path, one can see that 'starting from upper-layer scenarios and gradually building down capabilities' has been its consistent approach since inception.

Unlike Alibaba Cloud, Huawei Cloud, and Tencent Cloud, which 'first laid out IaaS data centers, then built PaaS platforms, and finally developed SaaS applications,' Volcano Engine originated as a cloud computing team serving ByteDance internally. When it officially commercialized in 2020, it first launched SaaS-layer products like Feishu, intelligent recommendations, and video cloud. It wasn't until September 2021 that it formally released IaaS infrastructure services like computing, storage, and networking, following a path of 'securing scenarios first, then building the foundation.'

In its early years, it entered through video cloud SaaS, gradually migrating entire business chains of small and medium-sized clients like Dongqiudi to its proprietary IaaS platform, validating this 'light entry, heavy retention' logic. Now, it aims to replicate this logic across the entire AI cloud market, a challenge orders of magnitude more difficult.

To judge whether Volcano's current moves represent mere external narratives or genuine strategic implementation, one need only observe whether each action continues to revolve around 'lowering barriers for new customer acquisition and expanding call volume' (a traffic-driven logic) or genuinely moves toward the 'tough business' of 'embedding into core processes and building industry trust.'

Cloud service barriers thicken layer by layer, but in real competition, it resembles an ever-deepening dive. The tool layer has the lowest barriers—anyone can do it. The middle application layer requires deep industry know-how. At the core lies trust, built through years of customer accumulation.

The outermost layer is tool capabilities. Products like Ark CLI essentially engineer model calls into standardized instructions, lowering development barriers, expanding the developer base, and increasing Token consumption. This leans more toward 'expanding traffic' than building barriers, as both leading cloud providers and MaaS platforms can offer similar capabilities, competing on efficiency and price with low developer switching costs.

Moving inward is the enterprise process layer. HiAgent 3.0 and the ArkClaw Enterprise Edition workstation begin integrating into real business processes: contract review, production workflows, and digital employee orchestration. The core here is not model capability but system integration capability: can it connect with OA and ERP systems? Can it adapt to industry regulations? Can it operate stably in production environments?

This 'delivery capability' offers no shortcuts—it requires industry-by-industry refinement and customized client adaptations, demanding extensive on-the-ground service teams and industry experience. Tan Dai mentioned in an exclusive interview that Volcano has now established FDE (Frontier Deployment Engineer) teams covering key industries like automotive, finance, manufacturing, and healthcare to implement scenarios jointly with clients.

This marks a departure from the 'selling Tokens and interfaces' light model, embracing the traditional cloud provider's 'heavy service, heavy delivery' path.

The deepest and toughest challenge is trust. The AI Trust system and its collaboration with China Mobile target critical sectors like government, finance, and energy. Data security, compliance, and stability represent impenetrable barriers—the same moats Alibaba Cloud and Huawei Cloud spent nearly a decade building through landmark projects.

Volcano's construction of a full-chain trusted computing system and its joint AI confidential computing zone with China Mobile aim to meet state-owned enterprises' core requirement of 'keeping data within boundaries.' However, it must be acknowledged that trust in such markets is never won through product launches or cooperation signings alone. From supplier qualification to the first pilot project and then large-scale core system procurement, the process often spans years, involving countless security tests and stability verifications.

Only by navigating this journey can Volcano Engine truly earn its ticket to the mainstream cloud market.

Conclusion

After more than a decade in China's cloud market, no provider has ever won solely through 'low pricing.'

AWS started by selling computing power for an online bookstore and waited a decade to become the global cloud leader. Alibaba Cloud built from Alibaba's own e-commerce systems, investing for ten years and enduring countless setbacks to secure its top position. Huawei Cloud leveraged three decades of to B customer relationships to break into the first tier within five years.

All cloud providers that ultimately survive and thrive must 'crunch the numbers' thoroughly. You can penetrate trial scenarios with extreme pricing, but to become a mainstream player, you must help clients account for adaptation and risk costs, ensuring they feel secure and get value when placing core businesses on your platform.

AI has given Volcano Engine a once-in-a-century opportunity to change tables, but while the table has changed, the rules for winning remain the same: price to 'pierce' costs, service to underwrite risks.

Editor: Muren Proofreader: Zhang Wenxin Production: Rui Zong

▼

Welcome to follow our video channel #I'm Dr. Qi#