ByteDance Raises Volcano Engine AI Revenue Target on Seedance 2.0 Demand
ByteDance’s Volcano Engine raised its full-year MaaS revenue target to RMB 15 billion after Seedance 2.0 became a larger AI revenue contributor. Seedance 2.0 is described as generating more than RMB 1 billion in monthly revenue, while average daily token consumption has grown by nearly 40% month-on-month. The practical question is whether Volcano Engine can keep video-generation usage converting into paid token consumption beyond high-usage content segments.
ByteDance Raises Volcano Engine AI Revenue Target as Seedance 2.0 Drives Paid Usage
ByteDance’s Volcano Engine raised its full-year model-as-a-service revenue target to RMB 15 billion, or USD 2.2 billion, after Seedance 2.0 became a larger contributor to its AI business.
The revised target is well above Volcano Engine’s full-year MaaS revenue of around RMB 1.5 billion in 2025.
Seedance 2.0 is now described as generating more than RMB 1 billion in monthly revenue, while a person close to Volcano Engine said average daily token consumption has continued to grow by nearly 40% month-on-month.
The commercial signal is that paid video-generation usage is becoming a more measurable revenue driver for ByteDance’s AI cloud business.
The practical question is whether Volcano Engine can keep that demand growing outside a few high-usage content segments.
Video Generation Carries the MaaS Upside
Seedance 2.0 was released on February 10 and quickly became the main growth engine in the account of Volcano Engine’s AI revenue plan.
Artificial Analysis Video Arena ranked the model first with a 1,269 Elo score.
Early demand also created queues during the Lunar New Year period, when generating videos slightly longer than ten seconds could take as long as ten hours.
The company then added access controls such as whitelists so annual-contract customers could use Seedance 2.0’s full-capability version more reliably.
That change marked the start of a clearer commercial path for the model.
In short dramas, Seedance has reportedly reached roughly 95% penetration.
That concentration makes the short-video segment an important test case for whether a video model can hold pricing power while text-chat model prices keep falling.
Coding Remains a Weaker Revenue Track
Volcano Engine’s growth is not evenly spread across every AI model category.
Coding has not brought much incremental revenue for Volcano Engine, with the source tying that weakness to ByteDance Seed’s coding capabilities.
Zhipu AI, also known as Z.ai, is presented as a stronger player in coding.
On an earnings call, Zhipu CEO Zhang Peng said GLM API call prices rose by about 83% in aggregate in the first quarter of 2026, even as API call volume grew 400%.
The contrast points to a split in China’s MaaS market.
Video generation still has enough differentiation to support paid demand for Seedance 2.0, while coding is more crowded and depends heavily on model performance, workflow loyalty and token economics.
The Next Revenue Test
Volcano Engine’s new target depends on whether Seedance 2.0 can keep converting usage into paid token consumption.
The model currently ranks second globally by market share behind Google’s Veo, and the stated goal is to expand overseas and become the world’s largest video model by market share.
Alibaba Cloud has set up several new MaaS sales teams since last year, and other major companies are also pushing coding plans and token-consumption strategies.
The next signal is whether ByteDance’s video-model momentum remains strong enough to offset weaker coding revenue and support another upward revision to Volcano Engine’s MaaS target.
