广告
加载中

与其盯着Seedance 2.5 不如先聊聊2.0 mini能否再来一个单月10亿?

壹叔团队 2026-06-29 13:46
壹叔团队 2026/06/29 13:46

邦小白快读

EN
全文速览

本文核心梳理了字节跳动旗下Seedance系列AI视频生成模型的商业化现状与最新动态,核心干货信息如下:

1. 已跑通的成熟业务数据:Seedance 2.0是字节目前确定性最高的AI生意,年化收入约143亿元,单月收入超10亿元,毛利率约70%,在短剧行业渗透率约95%,不靠补贴也不靠免费获客,直接向企业收费就能实现稳定盈利,这在国内AI商业化中十分少见。

2. 新品Seedance 2.0 mini的核心信息:该产品于6月上线,定位高性价比,价格比标准版下降一半,叠加优惠后720P单秒最低仅0.13元,支持480P/720P分辨率,保留多模态参考能力,已经开放第三方API接入。

3. 行业整体动向:目前普通消费者不会为AI生成工具持续付费,字节已经将AI重心转向企业服务,接下来行业将在电商营销AI视频领域展开竞争,新的增长不确定性较高。

本文分析了AI视频生成领域的商业化现状,对品牌商开展营销内容生产、把握消费趋势有较多参考干货,总结如下:

1. AI视频生成可帮助品牌大幅降低营销内容生产成本:低价版Seedance 2.0 mini生成720P视频最低仅0.13元/秒,价格比原版下降一半,同样预算可以产出一倍以上的内容变体,非常适合品牌生产社交短视频、广告切条、产品演示这类高频营销内容。

2. 行业与消费趋势:目前C端用户没有为AI内容生产工具持续付费的习惯,企业端是AI视频的主要付费市场,电商营销是除短剧外最大的增量市场,阿里、京东等巨头都已经在布局这块赛道。

3. 对品牌的价值:当前AI模型价格大幅下探,品牌商可以利用低价AI工具批量生产营销内容,降低内容生产门槛,匹配当下短视频、社交平台高频内容更新的需求,提升内容营销效率。

本文梳理了AI视频生成领域的最新变化,为从事AI内容相关业务的卖家整理了机会、风险与可参考方向,总结如下:

1. 行业最新格局:字节推出低价Seedance 2.0 mini并开放第三方API接入,原有短剧市场已经接近饱和,渗透率达95%,再想靠短剧挖掘新增量空间十分有限,接下来行业核心增长市场转向电商和营销领域。

2. 可把握的机会:低价AI模型作为底座开放API,第三方卖家可以将其接入自身业务工作流,面向电商、营销类客户提供AI视频生成服务,借助低价模型降低自身服务成本,拓展中小客户群体,打开新的增长空间。

3. 需要注意的风险:电商营销客户更看重服务的确定性、集成度和可控性,不会仅因为价格选择供应商,当前市场竞争激烈,阿里已经推出同类平台布局,京东也开源了相关框架,想要实现大规模增长难度很高。

本文介绍了AI视频生成领域的最新商业化进展,对从事内容生产相关的工厂把握需求、抓住商业机会、推进数字化有较多启示,总结如下:

1. 当前市场产品需求变化:短剧、电商、营销领域对批量AI视频生产的需求已经被市场验证,市场对高性价比、低成本的中分辨率视频需求旺盛,这类内容生产速度快、成本低,完全匹配当下市场高频内容更新的需求。

2. 可挖掘的商业机会:目前短剧市场已经饱和,但电商和营销领域的批量短视频生产还有很大增量空间,内容生产工厂可以对接成熟的低价AI模型,为品牌商、电商卖家提供批量短视频、广告切条、产品展示视频的生产服务,拓展新的业务增量。

3. 数字化转型启示:AI工具已经大幅降低内容生产的边际成本,传统内容生产工厂需要积极接入成熟AI大模型,改造自身生产流程,降低整体生产成本,提升生产产能,抓住AI带来的新市场机会。

本文分析了AI视频生成服务商赛道的最新发展趋势、客户痛点与可行方向,给相关服务商提供了不少干货参考,总结如下:

1. 行业发展趋势:AI视频生成目前已经跑通成熟商业化路径,To C市场无法实现盈利,企业服务是未来的主流发展方向,原有短剧市场已经接近饱和,电商和营销领域是接下来的核心增量市场,头部科技巨头都在加紧布局,行业竞争逐渐加剧。

2. 行业客户核心痛点:中小AI内容从业者成本压力大,大多找不到清晰的盈利路径,对AI模型降价有强烈需求;品牌、电商类客户需要低成本、高速度批量生产内容,原有高价模型的成本太高,严重限制了市场需求的释放。

3. 可行的解决方案:相关服务商可以依托头部厂商开放的低价AI模型底座,结合自身在应用层的服务能力,对接电商、营销客户的个性化需求,提供适配客户需求的一体化AI视频生产服务,抓住市场增量。

本文梳理了字节在AI视频生成领域的最新布局逻辑,给同类平台商的业务布局带来不少参考与风向提示,总结如下:

1. 企业端客户对AI视频平台的核心需求:存量高端市场饱和后,市场存在大量对高性价比中低端AI视频模型的需求,满足不同客户批量内容生产的降本需求,同时大量客户需要开放的API接口,方便将模型接入自身现有工作流。

2. 头部平台的最新布局做法:字节推出低价Seedance 2.0 mini,价格较原版下降一半,开放第三方API接入,依托自身剪映等应用矩阵铺量,以此争夺电商营销领域客户,应对阿里同类平台的市场竞争。

3. 需要规避的风向:To C端AI内容工具难以实现盈利,不要盲目投入大量资源在C端,核心增长点在企业服务;需要在原有市场饱和后及时拓展新的增量领域,通过分层产品覆盖不同客户群,竞争核心是应用层的确定性与集成能力,并非只有价格。

本文披露了字节Seedance AI视频生成模型的商业化数据与最新布局,展现了国内AI大模型商业化的最新动向,对相关领域研究有较高参考价值,总结如下:

1. 国内AI大模型商业化的新动向:国内AI大模型已经跑通可盈利的To B商业化路径,Seedance 2.0实现单月10亿元收入,毛利率约70%,短剧行业渗透率达95%,不靠补贴也不靠免费获客,直接向企业收费就能实现稳定增长,验证了AI大模型To B商业化的可行性。

2. 产业发展出现的新问题:头部模型原有核心市场已经接近饱和,算力供给长期紧张,边际收益不断下降;C端用户付费意愿极低,无法支撑大规模盈利,新的增量市场竞争激烈,对应用层服务能力要求更高,后续增长不确定性明显增强。

3. 新的商业模式参考:头部厂商推出分层产品策略,用低价轻量化模型开放API,覆盖更广的客户群体,依托自身生态矩阵争夺增量客户,为AI大模型商业化拓展提供了新的可研究模式。

返回默认

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

This article summarizes the commercialization status and latest developments of ByteDance's Seedance AI video generation models, with key takeaways as follows:

1. Mature business performance: Seedance 2.0 is ByteDance's most certain AI-powered business, with an annualized revenue of around 14.3 billion yuan, monthly revenue exceeding 1 billion yuan, and a gross margin of approximately 70%. It has achieved around 95% penetration in the short drama industry, and turns a stable profit by directly charging enterprise clients without relying on subsidies or free user acquisition — a rare case in China's AI industry.

2. Key details of the new product Seedance 2.0 mini: Launched in June, this model is positioned as a cost-effective option with half the price of the standard version. After discounts, it costs as low as 0.13 yuan per second for 720P resolution. It supports 480P/720P output, retains multi-modal reference capabilities, and has opened third-party API access.

3. Overall industry trend: Ordinary consumers are not willing to pay for AI generation tools on an ongoing basis, so ByteDance has shifted its AI focus to enterprise services. The industry will next see competition in AI video for e-commerce marketing, with high uncertainty around future growth.

This article analyzes the commercialization status of the AI video generation sector, offering actionable insights for brands to optimize marketing content production and adapt to consumer trends. Key takeaways are as follows:

1. AI video generation drastically cuts marketing content production costs: The budget-friendly Seedance 2.0 mini costs as low as 0.13 yuan per second for 720P output, half the price of the standard version. With the same budget, brands can produce more than twice as many content variations, making it ideal for high-frequency marketing content such as social short videos, ad cuts, and product demonstrations.

2. Industry and consumer trend: C-end users have not formed a habit of paying for AI content creation tools on an ongoing basis, so the enterprise segment is the main paying market for AI video. E-commerce marketing is the largest incremental market beyond short dramas, and major platforms including Alibaba and JD.com have already entered this track.

3. Implications for brands: With AI model prices dropping sharply, brands can leverage low-cost AI tools to produce marketing content at scale, lower entry barriers for content production, meet the demand for frequent content updates on short video and social platforms, and improve content marketing efficiency.

This article sorts out the latest developments in the AI video generation sector, and summarizes opportunities, risks and actionable directions for sellers engaged in AI content-related business:

1. Latest industry landscape: After ByteDance launched the low-cost Seedance 2.0 mini and opened third-party API access, the original short drama market has become nearly saturated with 95% penetration, leaving very limited room for new growth. The industry's core growth market will shift to e-commerce and marketing.

2. Available opportunities: As low-cost AI models open API access as an infrastructure layer, third-party sellers can integrate them into their own business workflows, offer AI video generation services for e-commerce and marketing clients, cut their own service costs with low-priced models, expand reach to small and medium-sized clients, and unlock new growth.

3. Key risks to note: E-commerce marketing clients value certainty, integration and controllability of services more than just price. The market is already intensely competitive: Alibaba has launched a similar platform, and JD.com has open-sourced related frameworks, making large-scale growth very difficult to achieve.

This article introduces the latest commercial progress in AI video generation, offering insights for content production factories to capture demand, seize business opportunities and advance digital transformation. Key takeaways are as follows:

1. Current changes in market demand: Demand for mass AI video production in short dramas, e-commerce and marketing has been market-verified. There is strong market demand for cost-effective, low-cost medium-resolution video; such content can be produced quickly at low cost, perfectly matching the current market's need for frequent content updates.

2. Exploitable business opportunities: While the short drama market is saturated, there is still large room for growth in mass short video production for e-commerce and marketing. Content production factories can connect to mature low-cost AI models to offer bulk production of short videos, ad cuts and product showcase videos for brands and e-commerce sellers, unlocking new business growth.

3. Insights for digital transformation: AI tools have drastically reduced the marginal cost of content production. Traditional content production factories need to actively integrate mature large AI models to revamp their production processes, lower overall production costs, expand production capacity, and seize the new market opportunities brought by AI.

This article analyzes the latest development trends, client pain points and viable paths for AI video generation service providers, offering valuable actionable insights:

1. Industry development trend: AI video generation has already built a mature, profitable commercial path. The C-end market cannot support profitability, so enterprise services are the mainstream direction going forward. The original short drama market is nearly saturated, e-commerce and marketing will be the core incremental market, and leading tech giants are all accelerating their布局, intensifying industry competition.

2. Core client pain points: Small and medium-sized AI content practitioners face heavy cost pressure, most lack a clear profit path, and have strong demand for further AI model price cuts; brands and e-commerce clients need to produce content in bulk at low cost and high speed, and the high cost of original high-priced models has severely restricted the release of market demand.

3. Viable solutions: Relevant service providers can build on the low-cost AI model infrastructure opened by leading vendors, combine it with their own application-layer service capabilities, meet the personalized needs of e-commerce and marketing clients, and provide integrated AI video production services tailored to client needs to capture market growth.

This article sorts out ByteDance's latest layout logic in the AI video generation sector, offering reference and trend insights for peer platform players. Key takeaways are as follows:

1. Core demands of enterprise clients for AI video platforms: After the high-end market becomes saturated, there is strong unmet demand for cost-effective entry- and mid-tier AI video models to meet different clients' need to cut costs for bulk content production. In addition, a large number of clients require open API access to integrate models into their existing workflows.

2. Latest layout moves by the leading platform: ByteDance launched the low-cost Seedance 2.0 mini, half the price of the standard version, opened third-party API access, and leverages its existing app ecosystem including CapCut to scale volume, in a move to capture clients in the e-commerce marketing space and compete with similar platforms from Alibaba.

3. Trends to avoid: C-end AI content tools are hard to monetize profitably, so platforms should not blindly invest massive resources in the C-end market; the core growth lies in enterprise services. Platforms need to proactively expand into new incremental fields after existing markets saturate, use tiered product lines to cover different client groups, and competition centers on application-layer certainty and integration capabilities, rather than price alone.

This article discloses commercial data and the latest layout of ByteDance's Seedance AI video generation model, reflecting the latest developments in commercialization of large AI models in China, offering high reference value for research in related fields. Key takeaways are as follows:

1. New developments in commercialization of large AI models in China: A profitable B2B commercialization path for large AI models has been proven in China. Seedance 2.0 generates over 1 billion yuan in monthly revenue with a 70% gross margin, holds 95% penetration in the short drama industry, and achieves stable growth by directly charging enterprise clients without subsidies or free user acquisition, verifying the feasibility of B2B commercialization for large AI models.

2. Emerging new issues in industrial development: The original core market of leading models is approaching saturation, long-term computing supply constraints have pushed marginal returns down; C-end users show extremely low willingness to pay and cannot support large-scale profitability, while the new incremental market is intensely competitive with higher requirements for application-layer service capabilities, leading to significantly higher uncertainty for future growth.

3. A new reference business model: Leading players are adopting a tiered product strategy, opening API access for low-cost lightweight models to cover a broader client base, and leveraging their own ecosystem to capture incremental clients, providing a new researchable model for the commercial expansion of large AI models.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

文/HAL

过去一年,Seedance 2.0大概是字节唯一具有确定性的AI生意。

根据《晚点LatePost》的说法,这款视频生成模型当前年化收入约20亿美元(合143亿元人民币),单月收入超过10亿元,毛利率约70%,绝大多数收入来自企业客户。在短剧行业,它的渗透率据称已达到约95%。

这是一组在国内AI商业化里很少见的数字。它既不靠补贴换规模,也不靠免费换用户,而是直接向企业收费,并且收得上来。

随之,增长的势头出现了变化。

多家报道提到,近期Seedance的收入增速有所放缓。几乎是同一时间,字节把AI资源的重心从豆包这类面向大众的产品,转向了面向企业的服务。

但是,豆包付费这件事能不能大获成功,业内多数的看法都是保留意见的,也因此,Seedance这把镰刀还需“越磨越锋利”,在此之下,6月中旬,它上线了一个更便宜的版本——Seedance 2.0 mini,而在刚刚结束的火山引擎Force大会上,Seedance 2.5也宣布于7月正式登场。

对于广大勤勤恳恳的AI创业者来说,Seedance 2.5华丽升级,也许不是痛点所在,他们更渴望的是当字节系Seedance成为绝对霸主,“降价”这件事能否有更多空间,毕竟之于大部分AI短剧、AIGC从业者来说,还没找到盈利的门道。

标准版模型升级,再秀肌肉,再搞炫酷炸天,不一定是增量的开始,而主动降价之下,是对于更多交付与落地的野望,其也许能带来更宽广市场的争夺,最终mini还能不能再跑出一条像2.0那样、单月10亿的曲线,或许现实情况远没有想的那么容易。

01

mini带来了什么

以及为什么是现在

6月16日,火山引擎旗下的火山方舟上线了Seedance-2.0-mini,定位是高性价比的视频生成模型。

价格是它最直接的变化。

不含视频输入的文/图生视频定价为0.023元/千tokens,含视频输入的视频生视频为0.014元/千tokens。按720P折算,单秒生成成本约0.5元,相比标准版Seedance 2.0约1元/秒,下降了大约一半。

如果叠加剪映的独家限时五折优惠,新人叠加包月首优最低可以来到720p 0.13元/秒,480p 0.08元/秒的“白菜价”。

规格上,mini支持480P和720P两档分辨率,时长4到15秒,统一24fps。它没有1080P,也就是说,它不是为高分辨率成片准备的。它保留了Seedance家族的多模态参考能力,最多可以接入12个参考输入。

mini首发于剪映,随后接入小云雀、即梦等字节系应用;面向开发者的API在6月22日开放,国内走火山引擎,海外走BytePlus。

API一旦放开,能调用它的就不止字节自家产品——大量第三方应用和Agent,也都可以把它接进自己的流程里。

它带来的好处很具体:单价砍掉一半,意味着同样的预算可以多跑一倍的版本和变体。对于社交短视频、广告切条、产品演示这类“质量门槛早已被跨过、真正的约束是速度和成本”的高频内容,这是实打实的降本。

相比于传言已久的2.1版本(现正式官宣为2.5版本),为什么mini先出来?

最直接的原因,是标准版的增量已经接近见顶。

短剧渗透率据称已达约95%,签了年框、走白名单的存量客户基本被吃透,再想靠短剧撑起一条新的增长曲线,空间已经不大。

供给侧也到了需要算账的时候,Seedance在春节期间出现过长时间排队,算力始终是硬约束。这两样都不是可以无限拉伸的,继续把资源压在高价、高算力消耗的业务上,边际收益只会越来越薄。

存量见顶,加上供给吃紧,剩下的出路自然是往更广的市场走,而要走出短剧,第一步就是把价格门槛降下来。mini的出现,本质上是为这件事做的准备——它不是为了在现有客户身上赚得更多,而是为了覆盖到更多客户。

02

更广的市场在哪里

以及mini如何再赚10亿

往更广的市场走,绕不开一个问题:谁来付费。

C端基本可以排除。豆包日活超过2亿,但每天收入不足百万元,而它每天消耗的算力成本高达数千万元。

这组对比已经说明,普通消费者不会为这类生成工具持续付费。

原因也不复杂——大多数人刷短视频是为了娱乐,不是为了生产内容;他们用AI视频工具的动机,远没有强到愿意为此付月费。字节把重心从豆包移向企业服务,本身就是一次用脚投票。

排除掉C端,剩下能规模化付费的,基本就只有电商和营销。这两个场景有明确的投入产出、有预算、也有批量生产的真实需求,是目前除短剧之外,唯一被验证能持续掏钱的AI视频市场。

而这块市场,阿里也已经开始布局。

5月底的阿里云峰会上,阿里云发布了AI视频创作平台“万镜一刻”,明确对准短漫剧和营销视频,提供工具平台、API和企业级套件三种方案,并已经接入了一批短剧、电商和营销领域的企业客户。

从这个角度看,mini的降价,与其理解成再造一个增长引擎,不如理解成和万镜一刻去争夺同一批电商营销客户——用更低的价格,尽可能多地覆盖这部分人群。

但这里有一个关键差别,决定了这场竞争不会靠价格本身分出胜负。

阿里的做法是直接端出一个一体化平台。字节的做法,是把mini当成低成本的底座,铺进自己的应用矩阵——先是首发的剪映,再到小云雀、即梦,并随着6月22日API的开放,进一步向更多第三方Agent敞开。

逻辑很简单:模型越便宜,愿意把它接进自己工作流的应用就越多。

所以,这场竞争真正的形态,更多是还是字节的Agent矩阵去向外四处竞争。

而如果指望靠这个降价的模型本身再跑出一个单月10亿,或许并不容易——便宜但低质的模型只是底座,它自己并不构成一门生意。

Seedance2.0mini真正的、也是唯一的机会,在于它接入的那些Agent能不能从五花八门的内容agent手里把电商和营销的工作流市场抢下来。毕竟就连京东最近也宣布开源了一个名为JoyAI-Echo的长音视频生成框架,主打长视频生成稳定性和连续性。

而这件事其实比让短剧客户买年框更难。

短剧是一个对模型能力断层领先高度敏感的市场,模型断层领先,客户不用也不行。电商和营销要的却是确定性、集成度和可控性,更需要应用层更更多确定性,不会单单只因为价格便宜就突然爆发额外需求。

下一个10亿如果还有,多半就在这场仗里。

只是它不会像第一个10亿来得那么快,更不会那么确定。

注:文/壹叔团队,文章来源:壹娱观察(公众号ID:yiyuguancha),本文为作者独立观点,不代表亿邦动力立场。

文章来源:壹娱观察

广告
微信
朋友圈

这么好看,分享一下?

朋友圈 分享

APP内打开

+1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0