首页  >>  来自播客: a16z Podcast
更新   反馈  

The Top 100 GenAI Products, Ranked and Explained

发布时间 2025-03-26 10:00:00    来源
A16Z 播客的这一集深入剖析了他们“Gen AI 100”榜单的最新排名，展示了基于月独立访问量和活跃用户数量排名前 50 的 AI 优先的网络产品和移动应用。由 A16Z 消费者合伙人 Olivia Moore 和普通合伙人 Anish Acharya 主持的讨论，着重强调了消费者 AI 领域的快速发展以及塑造这一领域的关键时刻。 分析的核心是 Gen AI 100 榜单本身，该榜单通过 SimilarWeb (用于网络) 和 Sensor Tower (用于移动) 的数据编制而成，分别根据月访问量和活跃用户数量对产品进行排名。本期榜单首次纳入了移动应用的收入数据，揭示了流量来源和实际盈利能力之间存在着显著的脱节。 对话首先概述了方法论，强调了其对数据的依赖，以便识别由 Gen AI 驱动的产品，即使它们没有明确地以此方式进行营销。这种方法旨在了解哪些 AI 应用真正引起了主流消费者的共鸣。随后，讨论转向了推动消费者 AI 采用的关键时刻。Midjourney 和 Character AI 在 ChatGPT 之前就已经存在，它们是消费者兴趣的早期指标。Balenciaga 教皇的图像被认为是唤醒公众对 AI 图像生成兴趣的文化转折点。进一步的动力包括 Snapchats 的 My AI、可口可乐的 AI 圣诞广告以及 BBL Drizzy 的歌曲。 嘉宾们强调了 AI 领域中不断打破假设的情况，指出每个关键时刻都驳斥了先前关于消费者行为、AI 能力或市场主导地位的既有观念。他们认为，许多直观的假设将被证明是错误的。他们强调，AI 视频终于开始奏效，并且 AI 正在降低创作成本。事实上，现在 95% 的 YC 公司都在使用这些工具。他们将 AI 采用的当前阶段定位在“早期采用者”阶段，将其与移动或云时代相提并论。他们还讨论了一些有趣的假设，强调了这些假设可能存在的问题。一种是人类将建立关系，而 AI 将擅长事务性互动。另一种是人类将工作委托给 AI，但 AI 可能将工作委托给人类。 讨论还深入研究了“准入榜单”，其中包括五家险些进入前 50 名的公司，包括 Runway、Order 和 U-Max。Cria 和 Lovable 等新来者正在持续上升。新来者的涌入强调了市场的动态性，挑战了停滞和老牌玩家的观念。AI 视频成为这些新来者中的一个重要趋势。该领域的重点公司包括 Hai Low、Clang 和 Sora。 对话还谈到了预期会出现但在榜单上没有出现的产品，例如更多的风格迁移和消费者语音产品。至于令人惊讶地出现在榜单上的产品，是氛围编码产品，因为它们被开发人员广泛使用。 某些公司在四个版本的榜单中持续出现，这突显了消费者 AI 领域中切实业务的发展。ChatGPT 作为此类应用开发的起点，仍然是顶级玩家。Deepseek 在榜单上也迅速崛起，即使当时只有 10 天的数据。 对话中也提到了留存率的问题。初步数据显示，DeepSeek 在用户参与度方面正在迅速赶上老牌玩家。Olivia 强调了其移动流量数据。在其他应用无法访问的国际市场，尤其是中国，留存率可能具有结构性优势。演讲者还讨论了目前哪些有效，哪些在产生收入。照片和视频生成器及编辑器，以及使用 ChatGPT 的美颜滤镜和克隆应用正在赚取大量资金并获得大量用户。在两个榜单之间只有 40% 的重叠。用户群越小，应用程序每次用户盈利的可能性就越大。最终，最好的产品是消费者认为有用且最令人愉快的产品。
The A16Z podcast episode dissects the latest rankings from their "Gen AI 100" list, showcasing the top 50 AI-first web products and mobile apps based on unique monthly visits and active users. The discussion, led by A16Z consumer partner Olivia Moore and General Partner Anish Acharya, highlights the rapid evolution of the consumer AI landscape and the pivotal moments that have shaped it. The core of the analysis is the Gen AI 100 list itself, compiled through data from SimilarWeb (for web) and Sensor Tower (for mobile), ranking products based on monthly visits and active users respectively. This edition marks the first time revenue data was incorporated for mobile apps, revealing a significant disconnect between what's generating traffic and what's actually making money. The conversation begins by outlining the methodology, emphasizing its reliance on data to identify Gen AI-powered products even if they aren't explicitly marketed as such. This approach aims to understand which AI applications are truly resonating with mainstream consumers. The discussion then shifts to pivotal moments that have fueled consumer AI adoption. Midjourney and Character AI, both pre-dating ChatGPT, are highlighted as early indicators of consumer interest. The Balenciaga Pope image is cited as a cultural turning point that awakened public interest in AI image generation. Further momentum includes Snapchats, My AI, Coke’s Christmas ad made with AI, and the BBL Drizzy song. The guests emphasize the constant breaking of assumptions in the AI space, pointing out that each pivotal moment debunked previously held beliefs about consumer behavior, AI capabilities, or market dominance. They argue that many intuitive assumptions will prove incorrect. They highlight that AI video is finally starting to work, and that AI is decreasing the cost of creation. In fact, 95% of YC companies are now using these tools. They place the current stage of AI adoption in the "early adopter" phase, comparing it to the mobile or cloud eras. They also discussed interesting assumptions, highlighting how they might be incorrect. One is the idea that the humans will build the relationships while AI will be good at transactional interactions. The other is that humans will delegate the work to AI, but AI might delegate the work to humans. The discussion also delves into the "brink list," comprising five companies that narrowly missed the top 50 cutoffs. These included runway, order, and U-Max. Newcomers like Cria and Lovable, which are on a consistent upswing. The influx of newcomers underscores the dynamic nature of the market, challenging the notion of stagnation and established players. AI video emerges as a significant trend among these newcomers. Highlighted companies in this space include Hai Low, Clang, and Sora. The conversation also touched on what was expected to appear on the list that was not there, which was more style transfer and consumer voice products. In regards to items that were surprisingly included on the list was vibe coding products, as they are being used widely by developers. The consistent presence of certain companies on the list over four iterations underscores the development of tangible businesses in consumer AI. ChatGPT, as the starting gun for this application development, remains a top player. Deepseek rose to prominence on the list as well, even with having only 10 days of data at the time. The topic of retention comes up in the conversation. Initial data suggests DeepSeek is rapidly catching up to established players in terms of user engagement. Olivia emphasized its mobile traffic data. Retention in international markets, particularly China, may have a structural advantage where other applications are inaccessible. The speakers also talk about what is currently working and what is generating revenue. Photo and video generators and editors as well as beauty enhancers and filters and clone applications that are using ChatGPT are making a lot of money and getting a lot of users. There was only 40% overlap between the two lists. The smaller the user base, the more likely the app was to make money on a per user basis. In the end, the best product is the product that consumers find utility and is most delightful.

摘要

This month, a16z’s Consumer team released the fourth edition of the GenAI 100 — a data-driven ranking of the top 50 AI-first web products and mobile apps, based on unique monthly visits and active users. In just six months, the consumer AI landscape has shifted dramatically. Some products surged ahead, others plateaued, and a few unexpected players reshaped the leaderboard entirely. In this episode, a16z General Partner Anish Acharya and Partner Olivia Moore join us to unpack the latest rankings and explore the key cultural and product moments that brought us to this point. Which applications are leading the pack — and which ones are quietly on the rise? What do trends like AI video, companion apps, and “vibe coding” reveal about the future of consumer AI? And for the first time, the team also analyzed which products aren’t just gaining users, but generating real revenue. If you’re looking to understand where we are in the GenAI adoption cycle — and what might come next — this episode offers a data-backed view into one of the fastest-moving corners of technology. You can find the full GenAI 100 list at a16z.com/genai100-4

GPT-4正在为你翻译摘要中......

中英文字稿  

Consumer activity typically lags by 6 to 9 to 12 months. What's happening on the research side? So many of these assumptions and that's why the assumptions they seem intuitively correct are going to turn out to be incorrect. We are finally on the verge of AI video starting to really work. It sort of follows the trend of AI decreasing the cost of creation in every way. 95% of YC companies are now building using those tools. I think compared to where we're going to be, we're still incredibly early.
消费者活动通常会滞后6到9到12个月。研究方面的情况如何呢？有很多假设看似直觉上是正确的，但最终可能会证明是错误的。我们终于处在AI视频真正开始奏效的边缘。这有点类似于AI在各个方面降低创作成本的趋势。现在95%的YC公司都在使用这些工具。我认为相比我们将达到的位置，现在仍然是非常早期的阶段。

This month our consumer team at A6CZ dropped our fourth installment of the Gen AI 100 list, a list of the top 50 AI first web products and mobile apps based on unique monthly visits and active users. And as our consumer team said themselves, in just six months the consumer AI landscape has been redrawn. Some products surged, others stalled, and a few unexpected players rewrote the leader board overnight. In today's episode we explored the latest rankings and the pivotal AI moments of the last few years.
本月，我们在A6CZ的消费团队发布了第四版《Gen AI 100》榜单，这是一个根据每月独立访问量和活跃用户数量评选出的前50个人工智能领先的网络产品和手机应用程序的榜单。正如我们的消费团队自己所说，仅仅六个月的时间，消费者AI的格局就被重新绘制。有些产品迅速崛起，有些停滞不前，还有一些意想不到的玩家在一夜之间改写了排行榜。在今天的节目中，我们探讨了最新的排名以及过去几年中关键的AI时刻。

Mid-journey and character AI both came out before ChatTubt. Remember Snapchats, My AI, the Balenciaga Pope, Coke did their Christmas ad. Each one of those unlocks broke down the assumptions that many of us held prior and have helped culminate hundreds if not thousands of AI applications that are now vying for our attention. The app store is going to be chaos. Yeah, I know. It's going to be chaos. So which applications top the charts this time around, whether household brand names or tools that you may have never heard of, plus where does this flurry of activity place us on the adoption curve?
在ChatTubt之前，Mid-journey和Character AI就已经面世了。还记得Snapchat的"My AI"功能、Balenciaga教皇的风潮，以及可口可乐的圣诞广告吗？这些事件每一个都打破了我们之前所持有的假设，并帮助催生了如今数以百计甚至是上千个争相吸引我们注意的AI应用程序。应用商店将会变得一片混乱。是的，我知道，那会很混乱。那么，这次究竟哪些应用会占据榜单的前列呢？不论是家喻户晓的品牌，还是你从未听说过的工具，以及这一系列活跃的表现让我们处于技术普及曲线的哪个位置？

And what trends did it out, like AI video or vibe coding, that give us a window into what's to come? Finally, this fourth edition of the list is actually the first time that we broke out what's actually making money. And today we have A16C consumer partner Olivia Moore and General Partner Anishacharya to break down all of the above. Of course, if you'd like to see the full list of the top 100 Gen AI apps, head on over to A16C.com slash Gen AI 100-4. Or you can check the link in our show notes. Okay, let's get started.
翻译成中文并表达意思如下： 有哪些趋势如人工智能视频或情感编码，给我们了解未来的机会？最后，这份榜单的第四版实际上是我们首次分析哪些项目真正赚到了钱。今天，我们邀请了A16C的消费者合伙人Olivia Moore和普通合伙人Anish Acharya来解析上述内容。当然，如果您想查看完整的前100名生成式AI应用程序列表，可以访问A16C.com网站的Gen AI 100-4页面。或者您也可以查看我们的节目注释中的链接。好了，我们开始吧。

As a reminder, the content here is for informational purposes only. Should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16C fund. Please note that A16C and its affiliates may also maintain investments in the company's discussed in this podcast. For more details, including a link to our investments, please see A16C.com slash Disclatures.
请注意，本文内容仅供参考，不应被视为法律、商业、税务或投资建议，也不应用于评估任何投资或证券，同时不针对任何A16C基金的投资者或潜在投资者。请注意，A16C及其关联公司可能也持有在本播客中讨论的公司的投资。欲了解更多详细信息，包括我们的投资链接，请访问A16C.com/Disclosures。

We're back for the fourth edition of the Gen AI 100 list. You guys have been working hard and tracking the consumer landscape for years now. But specifically for the last two and a half years since we really had that chat GPD moment. Tell me more about how you're tracking that ecosystem and how that comes through in this list. Yeah, it's super fun. This is one of my favorite reports that we've put together a couple times a year. We track the consumer AI landscape through what we do every day, which is like meeting with consumer AI startups that come to pitch us, seeing what goes viral on Twitter.
我们回来了，推出第四版的Gen AI 100榜单。你们一直以来都很努力，追踪消费者市场的动态，尤其是在过去两年半的时间里，自从我们有了那个Chat GPD时刻之后。聊聊你们如何追踪这个生态系统，并反映在这个榜单上的细节。是的，这真的很有趣。这是我一年中最喜欢撰写的报告之一。我们通过日常工作来追踪消费者AI的动态，比如与来我们这里展示的消费者AI初创公司会面，观察Twitter上哪些内容走红。

But actually there's a whole separate set of companies and products that might be reaching the true mainstream consumer. That might not even be marketing themselves as AI products, but they're powered by and made possible by AI. And so the whole original purpose of this report was to see how much overlap is there between those two categories what is the actual everyday person who might not know that they care about AI using in their day to day. That's great.
实际上，还有一整套不同的公司和产品，它们可能正逐渐为真正的主流消费者所接受。这些产品可能并没有以人工智能产品的名义进行营销，但实际上是由人工智能推动和实现的。因此，这份报告的初衷就是为了了解，这两类产品之间有多少重叠之处，以及那些可能并不知道自己关心人工智能的普通人在日常生活中实际在使用些什么。这很不错。

And so talk about the methodology, like what makes it onto this list or not? Because to your point, there's certain household names that you might see on Twitter or have that viral moment. But I think some people might be surprised to see what made it onto this list. So it's entirely based on data. We have two lists here, the top 50 on web and the top 50 on mobile. So the top 50 on web, we use a data provider called similar web, which tracks every single website globally.
那么，我们来讨论一下这个方法论，是什么因素决定了哪些内容可以上榜呢？正如你所说，有些家喻户晓的名字你可能在推特上见过，或者它们曾引起过关注。但我认为有些人可能会对这个榜单上的内容感到惊讶。主要是因为这完全是基于数据的分析。我们这里有两份榜单：一个是网络平台的前50名，另一个是移动平台的前50名。对于网络平台的前50名，我们使用了一个名为SimilarWeb的数据提供商，他们会追踪全球每一个网站的数据。

And we essentially go down and descending order of how many visits they get each month. For this report, it was January 2025. And then we go and we pick the first 50 of those that have the most monthly visits that are a Gen AI first products. We do something similar on mobile, but a different data set from sensor tower. For mobile, we look at monthly active users on the app. And then again, we pick the top 50 that are Gen AI products.
我们基本上按照他们每个月访问量的多少来进行排序。对于这份报告，这是2025年1月的数据。然后，我们挑选出访问量最多的前50个以生成式人工智能为主打的产品。对于移动端，我们采用了来自Sensor Tower的不同数据集，关注的是应用的月活跃用户。然后同样地，我们选出月活用户最多的前50个生成式人工智能产品。

And then for the first time ever, we actually looked at the top 50 on mobile by revenue, which we hadn't done before. And it was a really interesting experiment because the lists were pretty non-overlamping. Totally. And we'll get to that. Yes. We've been in this AI ecosystem for a few years now. In your eyes, what were the pivotal moments that led up to this point in time where we have, like you said, 50 on mobile, 50 on desktop, and a whole lot more in the wider ecosystem?
然后，这是我们第一次真正查看移动端收入前50的情况，这是我们之前没有做过的。这个实验非常有趣，因为这些榜单几乎没有重叠。完全不重叠。我们稍后会讨论这个。是的，我们已经在这个人工智能生态系统中待了几年。在你看来，在达到这一点之前有哪些关键时刻？就像你说的，现在我们有移动端的50，桌面端的50，以及更多广泛的生态系统中的内容。

You often say actually it's usually like the papers are written and then the models are developed and then applications are built on top of it. So the consumer activity typically lags by six to nine to 12 months, what's happening on the research side. So maybe just from the consumer awareness or behavioral perspective, there's a couple of moments for me. Actually mid-journey and character AI both came out before ChattGBT, which I think a lot of people don't know, but there was maybe these early niche communities of early adopters that were using both of those products in the summer and the fall of 2022 leading up to ChattGBT.
你常说，其实通常是先撰写论文，然后开发模型，接着再在其基础上构建应用。因此，消费者的活动通常会滞后研究进展6到9到12个月。所以从消费者意识或行为的角度来看，对于我来说，有几个时刻。其实，Midjourney和Character AI都在ChattGBT之前出现，我认为很多人不知道这一点。但在ChattGBT出现之前的2022年夏天和秋天，可能已经有一些小众社区的早期使用者在使用这两个产品。

And then post-ChattGBT things that just brought AI to consumer consciousness. So even remember Snapchats, my AI, with that little bot that appeared at the very top of your feet. And 150 million people used it. And for a lot of kind of younger consumers, that was actually probably their first real chance having a conversation with an LLM. On the image side, I think of the Valenciaga Pope, which was also, I think spring 2020, such a cultural moment. It was. And I think it made a lot of people realize for the first time that they should even be interested in AI images, because they could be that good and that convincing.
然后，在ChatGPT之后，很多事情让AI进入了消费者意识中。比如Snapchat的“我的AI”，那个出现在你界面顶部的小机器人。大约有1.5亿人使用过它。对于很多年轻消费者来说，这可能是他们第一次真正有机会和一个大型语言模型（LLM）进行对话。在图像方面，我会想到“Balenciaga教皇”，这也是在2020年春季的一个文化时刻。我认为这让很多人第一次意识到他们应该对AI图像感兴趣，因为这些图像可以做到这么好，这么逼真。

The first big AI music moment for me was, well, the BBL Drizzy song, which I think was a spring of 2024. And that also went mega viral. No book LM was another one. I think one of the moments where creative AI really shifted into almost enterprise consciousness was the end of last year when Coke did their Christmas ad. And a lot of that was generated by AI. And then of course the deep seek launch earlier this year. Deepseek was so interesting because I think it sort of had become settled wisdom that it would be very hard for horizontal model to get to mass consumer scale quickly again.
对我来说，第一次重大AI音乐时刻是BBL Drizzy的歌曲，我记得那应该是在2024年春天。这首歌非常火爆。此外，还有"无书LM"也是其中之一。我觉得创意AI真正进入企业意识的一个重要时刻是去年年底，当时可口可乐发布了他们的圣诞广告，其中很多内容都是由AI生成的。当然，今年早些时候的Deep Seek也引起了广泛关注。Deep Seek非常有趣，因为之前普遍认为，水平模型再次快速达到大规模消费者层面是非常困难的。

Like chat GPT had done it and chat GPT had become a verb and that opportunity had already been explored. And now we see deep seek growing as quickly as it did. And there's actually a couple of interesting nuances to deep seek. So one, I think important nuances, the fact that they released their reasoning model for free at scale. Previously you had to use a one pro and you had to pay chat GPT's premium subscription to get access to it. The other thing was just the product execution around chain of thought, which we've talked about a lot.
像Chat GPT已经成为一种动词一样，这种机会已经被探索过。现在我们看到Deep Seek也在迅速发展。关于Deep Seek，其实有一些有趣的细节。首先，其中一个重要的细节是，他们免费大规模地发布了他们的推理模型。之前，你必须使用Chat GPT的高级版本并支付订阅费才能访问它。另一个有趣的点是他们在产品执行方面对思维链的处理，这也是我们已经讨论过很多次的内容。

And I think is pretty well understood. But the fact that it showed you its thought process in real time was just super captivating. And now something that's become a step that every model takes. So I think it just really illustrates how early we are. You know, we as sophisticated users and investors are looking for further and further refinements. And once in a while something like deep seek comes out of the clear blue sky and just blows away all assumptions. Totally. And that word specifically assumptions. I think it's so key when you talk about these pivotal moments.
我认为这一点已经被很好地理解了。但事实上，它能实时展示思考过程，这真是令人着迷。现在这已经成为每个模型都要经历的一步。所以我认为这很好地说明了我们还处于一个非常早期的阶段。你知道，作为成熟的用户和投资者，我们在不断寻求进一步的改进。偶尔像"Deep Seek"这样的东西会突然出现，完全打破所有假设。完全没错。特别是“假设”这个词。我认为，当你谈论这些关键时刻时，这是非常重要的。

I feel like you could actually match each pivotal moment with an assumption like an assumption being oh well, I could never trick me into thinking a picture is real when it's not right. Right. Or I would never actually listen to a top 100 song that's generated by AI or chat GPT is cornered the market. No one else can penetrate it. Right. All of these assumptions that people are like, okay, sure. I was wrong about that prior one. But this one I'm pretty sure about.
我觉得你可以把每个关键时刻和一个假设对应起来，比如一个假设是“哦，我绝对不会被一张看起来像真的但其实不是的图片骗到，对吧？”或者“我永远不会去听一首由AI生成并进入前100名的歌曲”。又或者“聊天GPT已经垄断了市场，其他人都无法进入，对吧。”所有这些假设，人们可能会想，好的，我之前那个是错了，但这个我可以肯定。

We're seeing just like months. Yes. Being the delta between assumptions being broken. And so to your point on the arc of the market or the industry. I could see an argument where people are like, oh, we're actually pretty far along because we've already slashed all of those assumptions. But on the other hand, I'm hearing we still have a long way to go. So maybe put us along that arc. If we were to compare to the mobile era or the cloud era or previous technology areas.
在我们看来，这就像是几个月的时间。是的，成了假设被打破的一个差距。所以针对你关于市场或行业发展的观点，我可以理解有人会认为，我们已经走了很远，因为我们已经打破了所有那些假设。但另一方面，我也听到有人说我们还有很长的路要走。所以，也许可以将我们在这个发展轨迹上放一放。如果我们要与移动时代、云时代或以往的技术时代相比的话。

And are we in that early innovator stage still or are we somewhere else. I think we're still very much in the early adopter phase in many of these categories were just arguably still in the infrastructure building era and moving into the application building era. It depends on the modality like now LLM's are maybe people thought that was a solve problem. But then again, deep sea came in and upended all of that. There's a lot of things that are definitely not fully solved like AI video right now can generate great three or five or six second clips.
我们现在还处于早期创新者阶段吗？还是已经到了其他阶段？我认为，在很多领域，我们仍然处于早期采用者阶段。可以说，我们还在基础设施建设阶段，并逐渐进入应用程序开发阶段。这要看具体的模式，比如现在很多人认为大语言模型（LLM）已经解决了很多问题。但又出现了深度学习的进展，彻底改变了现状。还有很多问题尚未完全解决，比如现在的AI视频技术只能生成三到六秒的精彩片段。

But hopefully years from now we have AI video that can generate minutes long or even hours long movies. And so I think compared to where we're going to be. We're still incredibly early. Here are two assumptions that I think are interesting because it may turn out the reality is the exact opposite. One is that AI will be very good at transactional interactions, but humans will still be the ones to build relationships and connection. So an example of that would be what kind of phone calls are AI going to be best at.
希望在未来几年内，我们能够有可以生成数分钟甚至数小时电影的AI视频技术。所以，相比起未来我们可能达到的水平，目前我们还处于非常早期的阶段。这里有两个我认为有趣的假设，因为现实情况可能恰好相反。一个假设是，AI会非常擅长处理事务性互动，但是建立关系和连接仍然是人类的强项。一个例子是，AI可能最擅长处理哪种类型的电话。

And I think the assumption was well to be great at sort of scheduling and logistics and the exchanging of information facts. But we've heard over and over that in many cases the AIs are more human than humans. They just have more patients more nuance. They're never having a bad day. They're never hung over. So that's an interesting area of exploration. The other one that I think is interesting is the idea that humans will delegate work to the AIs and the AIs will do it. Like what if the AIs are the ones delegating the work to us. Perhaps AI is really good at organizing work and we're really good and also get a lot of joy out of doing it. So I think so many of these assumptions and that's why the assumptions they seem intuitively correct are going to turn out to be incorrect. Totally.
我认为，我们原以为AI在安排和协调日程、交换信息方面应该表现得很出色。但我们不断听到的是，在很多情况下，AI反而比人类更“有人性”。它们往往更有耐心、更为细腻，从不会情绪低落或宿醉。因此，这一领域很值得探索。另一个有趣的想法是，通常是人类将工作委托给AI来完成，那么如果AI反过来向我们分配工作呢？也许AI非常擅长组织工作，而我们则非常擅长并且乐于从事这些工作。因此，我觉得很多这些看似直观正确的假设可能最终会被证明是错误的。

And if we think about the report maybe one important data point is the fact that we see so many newcomers still right if we were in that later part of the innovation curve. You might expect more stagnancy might expect to see the same players, but every time you guys build out this report we're seeing all of these newcomers in this particular time the fourth report we saw 17 new companies on the web rankings in particular. And you actually have this quote where you say a few unexpected players rewrote the leaderboard overnight. So can you just speak to that in the movement that we're seeing.
如果我们考虑这份报告，其中一个重要的数据显示我们仍能看到许多新面孔。如果我们处在创新曲线的后期阶段，你可能会预期出现更多停滞，看到的是同样的企业。但每次你们完成这份报告时，我们都会看到很多新进入者。在这第四份报告中特别明显的是我们观察到了17家新公司出现在网络排名中。而且你还提到，有些意想不到的公司一夜之间改写了排行榜。你能对此变化谈谈你的看法吗？

One of the biggest trends among the newcomers is we are finally on the verge of AI video starting to really work not just for people who are enamored by AI and willing to generate a hundred times to get a good clip from people. Yeah, but for people who actually want to make something creatively in a condensed time period. So we had three new video models on the list this time. High low and clang which are both Chinese models and then Sora, which was open AI's model that was announced I guess more than a year ago at this point and finally was released. I think we'll see even more of a shake up here because VO 2 is the new Google model that is even next level beyond that from what we've seen in testing and that is probably finally going to hopefully come out in the next three or six months.
新人中最显著的发展趋势之一是，AI 视频技术终于要真正发挥作用了，不仅仅是那些对 AI 着迷并愿意尝试上百次来获得一个好片段的人可以使用这些技术，而是那些想在短时间内创造性地制作出作品的人也可以受益。因此，这次我们列出了三个新的视频模型：High、Low 和 Clang，它们都是中国的模型，还有 Sora，这是 OpenAI 的模型，据我猜测它早在一年多前就宣布了，终于现在发布了。我认为我们在这个领域会看到更多的变革，因为 VO 2 是谷歌的新模型，从测试中看来，它比前面提到的更先进，预计可能会在三到六个月内正式推出。

The other big category of newcomers were these vibe coding products cursor made the list that's more of like an agentic ID for a technical audience and then bolt made the list, which is for a non technical audience where you basically go from a text prompt to a fully functioning web app. Even though they made the list, I think we've still seen there's a really significant portion of their users that are people who are in tech and are actually technical, but they might be using something like a bolt or a lovable which made our brink list which we can talk about to maybe prototype something easier and then export the code and go and play with it themselves.
另一类重要的新产品是这些“氛围编码”工具。Cursor进入了该列表，这个工具更像是为技术受众设计的主动型IDE。而Bolt也上榜了，它是针对非技术受众的，你可以从一个文本提示生成一个完整的网络应用。虽然它们上了榜，但我认为我们仍然看到，它们的用户中有很大一部分其实是技术人员。他们可能使用Bolt或我们之后会讨论的Lovable这样的工具来更容易地进行原型设计，然后导出代码自行修改和实验。

So I think we haven't quite seen the vibe coding products hit the true mainstream user in terms of someone who's never worked in tech or developed an app. I love this category. It's so fun and it's so satisfying to actually see your ideas come to life in the case of bolt and lovable sometimes they are just sort of compelling interactive prototypes more than they are full fledged products, but that's usually enough to get a feel for whether this is something you want to invest deeper in. It sort of follows the trend of AI decrease in the cost of creation in every way and people just trying more ideas.
我觉得，"氛围编码"产品还没有真正走进大众用户的视野，尤其是那些从未在技术行业工作或开发过应用的人群。我非常喜欢这个类别，它非常有趣，当你看到自己的想法通过这类产品呈现出来时，非常令人满意。有时，这些产品更像是吸引人的互动原型，而不是成熟的产品，但通常这已经足够让人感受是否值得更深入地投入。这种趋势与人工智能降低创作成本的方向一致，人们不断尝试更多的想法。

I know just think about what that says about the untapped market of people who want to build this with code that this is on the top 50 list. And I think honestly both of them haven't had many apps built on them yet that have gone super viral. And though when that happens and I'm sure it will, those will become stories of their own, which will then increase awareness of the products of the true mainstream audience. I think we're going to see a really interesting diversity or range of products built on these which it might just be like this is my app that I just use for my very specific. Neesh pain point or there might be people who never learned how to code who want to build a venture scale product on something like a bolt or lovable and so seeing how that plays out will be very cool.
我知道，想想看，这说明了一个潜在的市场，有很多人想用代码来构建这个东西，以至于它进入了前50名单。我认为，老实说，这两者上面都还没有很多应用程序被开发出来并达到超级火爆的程度。不过，当这种情况发生时，我相信它会发生，这些应用程序本身就会成为故事，从而提高主流观众对这些产品的关注。我认为，我们将会看到在这些平台上开发出种类多样、非常有趣的产品。有些人可能会开发出满足自己非常特殊需求的应用程序，也可能有人从未学习过编程，却想在一个平台上构建一个具有商业规模潜力的产品，所以观察这一切如何发展将会很有意思。

Yeah, I think there's two phrases I've heard that I like one is sort of DIY or personal software but never made economic sense to design software for one really the other is disposable software just as soon known you do made a possible to make a song just to capture a joke that would be irrelevant the next day. These products make it possible to create a product or an experience that may have an extremely short shelf life like 20 minutes or a week or any other time period.
是的，我听过两个我比较喜欢的表达，一个是“自己动手”或者“个人软件”，但从经济的角度来看，设计这种软件从来没有意义。另一个是“一次性软件”，就像你写一首歌只是为了捕捉一个第二天可能就不再相关的笑话一样。这些产品使得创建一种产品或体验成为可能，即使它可能只有非常短的生命周期，比如20分钟或一周，或者任何其他时间段。

Let's talk about the brink list because that's completely new to this years for generation. So what is the brink list and why at it. So the brink list is essentially the five companies that almost made the list and were right below the cut off again purely based on the data. So we pulled the five websites in the five mobile apps. And I think honestly we were just curious to see what it would capture. We didn't quite know the takeaway for me.
让我们来谈谈“边缘榜单”，因为这是今年首次引入的新概念。那么什么是“边缘榜单”，为什么要设立它呢？“边缘榜单”实际上是指那些差点入榜而位于截止线之下的五家公司，这个排名是完全基于数据得出的。我们汇集了五大网站和五大手机应用。我想，我们只是出于好奇，想看看它会揭示什么。对我个人而言，我们并不完全清楚会有什么收获。

It does reflect how fast things are changing because there were a couple of companies on the list like runway, order, you max across web and mobile that have been on the core top 50 ranks in the past. But maybe they got just edged out by like deep seek watching this time and so they lost their spot for this ranking but might be on there the next one and they still massive usage.
这确实反映了变化的速度有多快，因为曾经有几家公司在过去的核心前50名榜单上，比如Runway、Order和UMax Across Web and Mobile。但这次它们可能被像Deep Seek Watching这样的公司挤出了位置，因此在这次排名中失去了席位，不过它们仍然有大量的用户使用，下次可能会重回榜单。

And then the other trend that it caught was a rise in more recent products like Cria made the list and lovable made the list that are very much on the kind of consistent upswing. And if it continues, we might see them on the main ranks and they haven't made the main ranks before.
翻译成中文易读版本：还有一个趋势是，它注意到了像Cria和lovable这样的新产品的崛起。它们很受欢迎，并且一直在稳步上升中。如果这种趋势持续下去，我们可能会在主排行榜上见到它们，因为它们之前从未上过主排行榜。

What did you predict that you would see on the list that you didn't really see there were there any surprises on that end. So one thing I thought we'd see more of is style transfer as an approach to scalable video because style transfer is just a much more tractable problem and has a lot lower cost of inference versus raw text to video.
你预期会在名单上看到哪些内容，但实际上没有看到？在这方面有什么让你感到惊讶的吗？有一点我之前以为会看到更多的，就是将风格迁移作为一种可扩展视频的方法。因为风格迁移问题更易于处理，相较于直接将文本转化为视频，它的推理成本要低得多。

But researchers and product developers seem to be really going for it on text to video and we've seen more of that than I would have expected. The other things that we didn't see on this list that we have seen at the model level. So that means maybe they'll be on the next list or like consumer voice products. There are a few of them, but not a ton of them. Some of the new like the Gemini flash model that can see what's going on on your screen and interact with you. Like I built something to yell at me if I go on Netflix or something. Like it's like, it's the best detector. Yeah, it starts like screaming. I you like know get back to work. Or like the new open AI operator model, which can actually interact with things on the browser level on your computer and get tasks done for you like pay a bill or make a graphic design or hire someone to landscape your yard something like that.
研究人员和产品开发者似乎正在全力推进文字转视频技术，这种进展比我预期看到的还要多。我们在这个清单上没有看到的其他东西，已经在模型层面上出现了。这意味着它们可能会出现在下一个清单上，比如消费类语音产品。这类产品只有少数几款，并不多。有些新技术，比如Gemini闪光模型，可以监测你屏幕上的活动并与你互动。我还做过一个类似的东西，它会在我打开Netflix时冲我大喊，提醒我回去工作，就像一个最佳的检测器。或者像新推出的OpenAI操作员模型，可以在电脑浏览器层面与你进行互动，帮助你完成任务，比如支付账单、制作图形设计，或者雇人整理庭院等等。

I think there's always a lag because the models have to be released to developers and they have to be tuned by the developers and so takes a while. But I would expect to see maybe an explosion of fun and unique and interesting products built on models like that on hopefully the next list or two because it feels like we really have seen an explosion on the model side and it is right there in terms of manifesting at the app level too. So one of the examples of this is deep research, which if you played with it is completely magical. But it's a primitive right it's not a product. It's something to build other things with. So it's really unclear if deep research is going to be used to write college thesis or is it going to be used to find the perfect meme to match a joke you want to make.
我认为总会有一些延迟，因为这些模型需要发布给开发人员，并由他们进行调优，所以需要一些时间。不过，我预计在接下来的一个或两个列表中，可能会看到基于这些模型的有趣和独特的产品的大量涌现。因为在模型方面，我们已经看到了爆炸式的增长，而且这种增长也即将在应用层面显现。其中一个例子是深度研究，如果你尝试过，就会觉得它如同魔法。不过，它更像是一种基础工具，而不是一个最终产品。它可以用来构建其他东西。因此，目前还不清楚深度研究是会被用于撰写大学论文，还是会被用来寻找完美的表情包来搭配你想说的笑话。

And that's all going to be up to the app developers and just a double click on that because you could see maybe a world where deep research is just this like more broad horizontal application or you could see what you just described. Or developers are tailoring that to specific and use cases. Are you basically saying that you think the latter is more likely in terms of the progression of these models and apps not more likely. But I think it's under explored. If you come to deep research today, you have the blank page problem. And I'd love to see developers create some constraints that leads to unexpected outcomes. Yeah, like the known or the prescribed use of deep research right now is basically market research reports and it's amazing for that. I've used it for that a lot of times.
这完全取决于应用程序开发者，让我再详细说明一下。你可以想象一种情况，深度研究是一种更广泛的通用应用程序；也可以像你所描述的，开发者将其定制化，用于特定的使用场景。你是否在说，你认为后者更可能是这些模型和应用发展的方向？我并不认为它更可能，但我认为这是一个未被充分探索的领域。如果你今天接触深度研究，你会遇到空白页问题。我希望开发者能设置一些约束条件，从而产生意想不到的结果。是的，目前深度研究的常见或指定用途基本上是市场调研报告，对此它非常出色。我自己也多次用它进行市场调研。

But if you try other things like one day we were trying to trace the origin of a meme and deep research is like a hundred expeter version of that. Know your meme website that kind of goes through the history. And the etymology or however you describe it. That comes to me. I mean, that should be an app. Yeah. So there's lots of other use cases that aren't market research reports that could really benefit from an incredibly obsessive compelling model that will go and read every website on the internet until it finds the answer. I love that I've actually always wanted that for creators because you know how there's the whole success overnight phenomena that everyone else thinks happens, but it's not true for most creators who are like you take Mr. Beasties. Like I literally counted to what 100,000. Yeah. And then I did thousands of more videos until I like something started to work.
但如果你尝试其他事情，比如有一天我们试图追溯一个表情包的起源，而深入研究就像是这种尝试的一百倍升级版。就像那个“了解你的表情包”网站，它会梳理其历史和词源，或者你怎么描述都好。我觉得这应该是一个应用程序。对，还有很多其他使用场景，不仅仅是市场调研报告，可以真的受益于一个极其执着、引人入胜的模型，这个模型会不停地浏览互联网上的所有网站，直到找到答案。我一直希望为创作者提供这样的工具，因为你知道，很多人觉得“成功一夜之间”现象是真的，但对大多数创作者来说并不是这样。比如说Mr. Beast，他曾字面意义上数到十万，然后制作了成千上万的视频，直到有些东西开始奏效。

And I wish you could actually just see as you're saying history of a meme, but history of a creator like when did the unlock happen. So those are the things that you thought might be on the list, but you didn't actually see there. What about the opposite. I think the fact that the vibe coding products like the bolt and the cursors and loveables made the mainstream consumer list is just a testament to how widely they're used by the technical audience like they have gotten to saturation so quickly. I think Gary Tan had some tweet that like 95% of YC companies or something are now building using those tools with something that nearly every developer now is probably using, which was maybe a surprise to me how quickly we reach saturation.
我希望你能看到的不是一个网络迷因的历史，而是创作者的历史，比如什么时候出现了突破。这些是你以为会列出的内容，但实际上并没有看到。那么相反的又是什么呢？我认为那种像 bolt、光标和 loveables 这样的编码产品出现在主流消费者列表上，证明了它们在技术受众中的广泛使用，它们如此快速地达到饱和状态。这让我想起 Gary Tan 发表的一条推文，说大约 95% 的初创公司现在都在使用这些工具。几乎每个开发者可能都在用，这让我对它如此迅速地达到饱和感到有些惊讶。

We've talked about this, but a continuing surprise. So I don't know if it counts as one, but I still have surprised every time is how many companion products are on the list and also how many of them rank so high. I think we had three companion products in the top 10. Two of them were NSFW oriented. Maybe not surprising when you think about like traffic on the internet in general outside of AI, but a lot of people are even using them as like interactive sandfiction. And some of the biggest fanfiction sites in the world are also top 100 top 200 global sites. So it makes sense in that way.
我们已经讨论过这个问题，但它仍然让我感到惊讶。我不知道这算不算是惊讶，但每次看到有这么多相关产品在榜单上时，我仍然感到意外。我们在前十名中有三个相关产品，其中两个是成人向的。考虑到互联网整体流量情况，这也许并不令人意外，不过很多人甚至把它们当作互动小说来使用。而世界上一些最大型的同人小说网站也在全球前100或前200站点之列，所以从这个角度来看，这也是合情合理的。

And then I guess my last surprise would be there's actually quite a bit of consistency in the list over the past four versions. There's always new entrance, which is really exciting, but across the four list, there's now 16 companies on the web ranks who have made it every single time and have kept the street going, which is pretty remarkable when you think of how early we are. But I think a testament to how those companies have cemented their brands, their products, their kind of I guess status and consumer consciousness. And I think a testament to the fact that like real businesses have been built in consumer AI already.
然后，我想我最后一个感到意外的地方是，在过去的四个版本中，这个名单实际上相当一致。总有新的企业加入，这真的很令人兴奋，但在这四个名单中，有16家公司每次都榜上有名，并且一直保持着这个趋势。考虑到我们目前这个领域还处于早期阶段，这一点相当了不起。这也证明了这些公司已经牢固地建立起了他们的品牌、产品和在消费者心目中的地位。我认为这也证明了在消费者人工智能领域，真正的业务已经建立起来了。

No, to add to that, one of the surprises for me on companion was not seeing more multimodality. Yes. The kind of the first glimmers of that at scale were GROC. GROC added a bunch of voices with some real aesthetics and points of view. It's a good way to put it. But it's just interesting that that feels in it. Of course, characters got voice mode and more, but it feels like character is in companionship is such a horizontal category. There's so much latent demand and a really increase once you have multimodality.
不，我想补充一点，对我来说，伴侣（或者说伙伴关系）中一个令人惊讶的地方是，没有看到更多的多模态性。是的。实际上，最早在大规模中看到这一点的迹象是从GROC开始的。GROC加入了许多带有真实美学和观点的声音。这是一种很好的描述方式。但有趣的是，这种感觉就是在其中。当然，角色有了语音模式和更多功能，但感觉角色在伴侣关系中是一个非常横向的类别。隐藏的需求非常大，一旦你有了多模态性，这种需求就会大幅增加。

The other interesting thing is that a lot of the text to code work, my assumption was that there was a small number of people who are creating sites that were heavily traffic and that explained the rise of them. But actually the majority of the traffic correct me if I'm wrong here, Olivia, is from people doing creation, not just consuming other people's creations. So there's just it really shows how much demand there is to make things. Even if people are not that interested in consuming them. You can track the traffic of apps that people have launched on lovable.app versus visits to lovable.dev, which is where people go to make lovable products. And lovable.dev has more usage or visits significantly than traffic to lovable.app, which gets back to what I was saying before.
另一件有趣的事情是，我原以为只有少数人在创建受欢迎的网站，导致其流量上升。但实际上，大部分流量，Olivia，如果我说错了请纠正我，来自于人们进行创作，而不仅仅是消费他人的作品。这确实显示出人们对创作事物的需求有多么强烈，即使他们对消费这些作品没有太大兴趣。你可以比较人们在lovable.app上发布的应用程序的流量与访问lovable.dev的次数，后者是人们创建可爱的产品的地方。而lovable.dev的使用或访问明显多于lovable.app上的流量，这正好印证了我之前所说的。

We have not even seen the first wave of viral products built on top of lovable and bold. And so when that happens, I think the awareness of these types of platforms is going to go significantly up show. The app store is going to be chaos. Yeah. It is going to be chaos. We're going to need an AI just to solve that AI app management problem. Completely. To that end, you talked about the fact that there are some consistent players.
我们甚至还没有看到那些基于“可爱”和“大胆”构建的病毒式产品的第一波浪潮。因此，当这种情况发生时，我认为人们对这些平台的认知会大大提高。应用商店将变得一片混乱，对，真会是一片混乱。我们将需要人工智能来解决AI应用管理的问题。确实如此。对此，您提到了有一些在这个领域中持续活跃的参与者。

Yeah. One of those players is Chatchy BT, which we've talked about as the starting gun of some of this application development. Chatchy BT has been at the very top of the list. Has it been that way for a time, single iteration of web and mobile? But maybe what would surprise people is that the traffic to Chatchy BT hasn't always been the same trajectories. Maybe can you talk about that and what did we see this time around? So it was basically flat for a while, which I think was surprising to a lot of people between February, 2023. Basically for a whole year through February, 2024, it was essentially flat in monthly visits to the website.
好的。其中一位玩家是 Chatchy BT，我们之前将其视为某些应用开发的起点。Chatchy BT 一直名列前茅，但也许令人惊讶的是，Chatchy BT 的流量并不总是沿着同样的轨迹增长。能否谈谈这一点，我们这次看到了什么情况？实际上，在 2023 年 2 月到 2024 年 2 月之间，也就是大约一整年的时间里，访问网站的月流量基本是持平的，这让很多人感到惊讶。

And I think at that point, from the date of it, I've seen basically 50% plus of the traffic with students who are using it for essays or homework problems. But the vast majority of other people, me included, to be honest, had not maybe found a daily active use case for Chatchy BT yet. And it's completely researched more recently. So they two X to the number of visits on web since then. They actually made their own announcement to where they counted across web and mobile. And in the past six months, they grew from 200 million to 400 million weekly active users. Wow.
我认为从那时起，根据其日期，基本上超过50%的流量来自于使用它撰写论文或做作业的学生。不过，大多数其他人，包括我自己在内，说实话，可能还没有找到一个每天活跃使用Chatchy BT的理由。但最近，他们实现了彻底的重新研究。他们的网页访问量翻了一番。事实上，他们自己也发布了公告，统计了网络和移动端的数据。在过去的六个月中，他们的每周活跃用户从2亿增长到了4亿。哇。

Which is especially surprising because it took them nine months to double before that. And it usually gets way harder to double at scale, not easier. I think from our perspective, if you've even plotted on the graph, you can kind of track the increases to the release of new models that unlock new use cases. So like the new 01 reasoning models, the 4-O models, which were multimodal for the first time and then advanced voice mode. And then they've also launched new products like the operator that can perform tasks in your computer like Canvas where you can write more naturally.
这尤其令人惊讶，因为他们之前花了九个月才实现规模翻倍。通常来说，规模越大，翻倍越困难，而不是简单。在我们看来，如果你在图表上标注，你可以追踪到每次的增长都是因为新模型的发布，这些模型解锁了新的使用场景。比如，新的01推理模型，4-O模型（首次实现多模态功能），以及先进的语音模式。 他们还推出了可以在你电脑上执行任务的新产品，比如Operator，可以在Canvas上更自然地书写。

So it's both bringing in new users who never tried it and then taking people like me who honestly I was maybe a weekly, if not less, a weekly active use case. And now I'm a daily active, but across several use cases now. Some days I'm driving and talking to it voice mode. Some days I'm working on a memo and I'm generating something with the research. Some days I'm doing some random other project and I'm brainstorming ideas with it. So I would expect that to continue as they release new models.
这不仅吸引了从未尝试过的新用户，还把像我这样的人转变了过来。老实说，我之前可能每周才用一次，甚至更少，但现在我几乎每天都会用它，而且用途多种多样。有时候我在开车，使用语音模式和它交流。有时候我在写备忘录，用它来做一些研究生成。有时候我在做一些其他随机项目，用它来头脑风暴一些想法。我预计随着他们推出新模型，这种情况会持续下去。

Have you heard from the ecosystem in terms of what more frequent use cases have emerged like yours in terms of if before it was a lot of students writing research reports? Is there a sense of understanding of what those newer use cases are? Yeah, I think it's gotten better at some things related to coding. It's gotten better at data analysis. And then I mean, the reasoning models, it's hard to overestimate because in the past, you couldn't even rely on Chachy. We need to tell you how many hours were in strawberry accurately. So it's hard to feel good about really tasking any sort of delicate or serious work to it. And so I think there are probably a long tail of use cases that people have just migrated over and out of they have more competence in the models.
你有没有从生态系统那里听说过，像你这样的应用案例是否变得更频繁？以前是不是有很多学生在写研究报告？对于那些新的应用案例，有没有更清晰的理解？我认为，现在在编程和数据分析方面的能力有所提高。而且，在推理模型方面的进步是不可低估的，因为过去你甚至无法依靠像Chachy这样的工具来准确告诉你草莓中有多少小时。因此，要让它承担任何细致或重要的工作，感觉很困难。我觉得可能有很多用户逐渐把这些应用案例迁移过来，因为他们对模型的信心提升了。

What's interesting to add to that is that Quad is not a traditional number two player. Typically the number two player has 10% of the market share and 10% of the product quality. And instead Quad sits in this very interesting place where it seems like it's more beloved by a smaller number of people. It's better at creative writing. It seems to have more of a personality, which is interesting because at least I think it's designed to be more constrained. And then it's also strangely much much better at coding. Yes. Why? I don't know. But it's very interesting to see there's a place for both Chachy, Pt and Quad and Mistral and potentially other models all to sort of augment each other.
有意思的是，Quad并不是传统意义上的第二名玩家。通常来说，市场排名第二的公司拥有10%的市场份额和10%的产品质量。然而，Quad处在一个非常有趣的位置，它似乎更受一小部分人的喜爱。它在创意写作方面表现得更好，似乎更有个性，这很有趣，因为我认为它原本是设计得更受限的。而且奇怪的是，它在编程方面要好得多。为什么会这样，我不知道。但很有趣的是，Chachy、Pt和Quad还有Mistral，以及可能的其他模型，都在这个领域中各自补充着彼此。

The really interesting thing about this list when it came to general LLM assistant usage was like, we only had 10 days of data for deep seek for January because it launched at the end of the month. And it shot up from literally nothing to number two on the list, 10% of Chachy, Pt scale on web within a week, a little bit more than a week. On mobile, it had even less than that, five days. And it was number 14 and if it had five more days, it would have been number two. And the gap is even narrower there between deep seek and Chachy, Pt. So again, to an issue's point, like that was a surprise and that we could see kind of a broad base sell product go so viral still and capture so many users. And deep seek was obviously the story when it came out.
这份清单中关于通用大型语言模型助手使用的一个非常有趣的点是，我们对于一月份的Deep Seek只有10天的数据，因为它是在月底推出的。它的使用量从无到有迅速上升，直接跃居榜单第二位，占到了Chachy Pt网页版本使用量的10%，而这仅仅在一周多一点的时间里。在移动端，它的表现更惊人，只有5天的数据，就已经排到了第14位。如果再有五天时间，它可能会成为第二。Deep Seek与Chachy Pt之间的差距在移动端甚至更小。因此，正如某个问题提出的那样，这是个意外，因为我们看到这样一个面向大众的产品仍然能够如此迅速地走红并吸引大量用户。Deep Seek显然成为了发布时的一个焦点。

What have we learned about retention since then? And is that learning specific to deep seek? Yeah. Or are we seeing that learning applied across the ecosystem? It's a little early call and retention and also because they're giving away so much for free right now. It's somewhat easy to retain. I will say the mobile data is fairly conclusive. So you can essentially look at sessions per week and time per week for any app. So we looked at perplexity, claw, deep seek and Chachy, Pt. Deep seek is already at the levels of perplexity and claw, which is interesting. So users are spending about 20 minutes a week across 10 sessions.
自那时以来，我们在用户留存方面学到了些什么？这种学习是否特定于Deep Seek？或者说，我们是否看到这种学习已经应用在整个生态系统中？因为目前他们免费提供了很多东西，所以要全面评估留存率还有点早。也因此，现在留住用户相对较容易。但我可以说，移动数据非常具有说服力。你可以查看任何应用程序的每周会话次数和每周使用时间。我们查看了Perplexity、Claw、Deep Seek和Chachy Pt的数据。令人有趣的是，Deep Seek的使用水平已经与Perplexity和Claw持平。因此，用户平均每周在10次会话中花费大约20分钟。

Still pretty significantly lags Chachy, Pt, which has like 45 minutes a week. So it's already at the level, if not actually slightly better. And this is a chart we put in the report to versus perplexity and clawed in terms of engagement. On a retention basis, like how many users are coming back to the app, say exactly 30 days, exactly seven days, exactly 60 days. It's just slightly below Chachy, Pt. So we're looking at 7% day 30 for deep seek and 9% day 30 for Chachy, Pt. It's too early to call on web because it's hard to track usage. Part of my theory here is if you look at deep seek usage, a lot of it is the US, but a lot of it is China and other countries where you can't use it.
仍然显著落后于ChatGPT，后者每周的使用时间大约是45分钟。所以尽管水平已经相当，甚至可能稍微更好一些，这是我们在报告中用图表展示的，以与复杂性和Claude在用户参与度方面进行对比。从用户留存率来看，比如说，恰好30天、7天、60天有多少用户会再次使用该应用程序，数据显示比ChatGPT稍低一些。具体来说，Deep Seek在第30天的留存率为7%，而ChatGPT是9%。在网页端的使用情况上，目前还很难判断，因为难以跟踪使用情况。我的一个理论是，如果你查看Deep Seek的使用数据，会发现很多用户在美国，但也有很多是在中国和其他无法使用的国家。

Or they try to make you not use it and you can only get by it with a VPN. And so in those markets, it's not Chachy, Pt versus deep seek versus perplexity. It's deep seek versus nothing. And so in those markets, I think they have like a structural advantage from the retention side that might skew the overall sample. Totally. Next time we should add a different cut for deep seek USA. Yes, exactly. Yeah, geographic breakdown. Yeah. Talking about trends that we're seeing on the list, you mentioned AI video before, but anything else you want to call out there in terms of its presence on the report?
或者，他们会尝试让你不能使用它，你只能通过VPN绕过。在这些市场里，不是Chachy、Pt与Deep Seek及Perplexity的对比，而是Deep Seek与什么都没有的对比。因此，我认为在这些市场中，Deep Seek在用户留存方面具有结构性优势，这可能会影响整体样本。完全同意。下次我们应该为Deep Seek美国市场增加一个不同的分类。对，正是如此。是的，地理分类。说到我们在报告中的趋势分析，你之前提到AI视频，但还有其他什么想要特别指出的吗？

Two of the video models were Chinese video models, which is super interesting. The models are less copyright sensitive in their training data. It's a great euphemism. Yeah, there are maybe more realistic and more prompted here and in the outputs as a result. But also just in China, it's easier to hire people to kind of caption videos. They have maybe a greater volume of researchers doing image and video staff versus other stuff. I think Sora in some ways was a little bit disappointing for some people, whereas like the Chinese video models were maybe better than a lot of people expected, given the relative lack of capital that they've raised.
这段文字翻译成中文大意如下： 两款视频模型是中国的视频模型，这点非常有趣。这些模型在训练数据方面对版权的敏感度较低，这是一个很妙的委婉说法。是的，可能因此这些模型在这里以及在输出结果中显得更为真实和受提示。但在中国，雇人给视频加字幕相对容易一些，他们可能有更多的研究人员在从事图像和视频方面的工作。有人认为Sora在某些方面让人有些失望，而相较之下，考虑到中国视频模型在筹集资本方面相对较少，这些模型的表现可能比很多人预期的要出色。

That's right. I think an interesting trend is just seeing Korea and the brink list. Yes. Korea is the single best place to access all the models and all the tools. And the nice thing that they do is stitch all of these things together to make them greater than some of their parts. So insofar as we live in this sort of multi-polar world of models, image models, video models, language models, they'll be a role for aggregators like Korea to put the model together in a potful way. Totally. Especially because people who are deep in AI video understand this, but each model is known for being good at specific things, like shots of people, shots of landscapes, anime, hyper realistic. And so it can rack up very quickly on $20 a month's subscriptions if you're paying for 10 or 15 different models independently versus having one canvas to work with all of them. You also typically use the products together.
没错。我认为一个有趣的趋势是看到韩国在这个领域的突出表现。是的，韩国是获取所有模型和所有工具的最佳地点。令人欣喜的是，他们把所有这些东西整合在一起，使其发挥出比各自独立时更大的作用。在我们生活在这个多极世界中，有图像模型、视频模型和语言模型，像韩国这样的聚合器将会在把这些模型有效组合在一起方面发挥作用。尤其是在AI视频领域深入了解的人知道，每个模型擅长的都是特定方面，比如人物镜头、风景镜头、动漫、超现实主义等。如果你独立购买10到15个不同的模型，每月20美元的订阅费用会很快累积，而有一个平台可以综合使用所有这些模型就更为划算。此外，你通常也会把这些产品结合起来使用。

Yeah. Usually generate an image in mid-June or flux and then you take that image and upscale it and then you put it as the beginning frame in a video. So you really want to not have the seams between all those products. So, you can see that in the video model, you can see that the image models are completely, are we seeing these video models in particular become more opinionated? And what I mean by that is we see that in image models, right, where mid-June might be good at this and then you might see another model better at something else. The users will gravitate towards either models or applications that provide them with that specificity or a piece. Are we seeing that in video?
好的。通常我们在六月中旬或在flux中生成一个图像，然后将这个图像进行放大，并将其作为视频的开头画面。因此，你会希望在这些产品之间不要有明显的缝隙。在视频模型中，你可以看到图像模型的效果，而且我们特别注意到这些视频模型变得更加有针对性。我的意思是，在图像模型中，我们可以观察到六月中旬时的某些模型擅长于此，而另一些模型可能在其他方面表现更佳。用户会倾向于选择那些能提供他们所需特性或细节的模型或应用。那么在视频中，我们是否也看到了这种趋势呢？

Yeah. So, you can see that in the video model level, but also the application choices that they're making that even the model companies are making are becoming more opinionated. If you've used like a runway or a clean or something, you can now prompt basically the camera angles or the wideness of the shot or all of these things. A human cinema photographer would do. You can prompt how the video sweeps over the surface of the screen. And so that's also the big factor in what you use for maybe even different parts of one video, which is interesting.
好的。你可以在视频模型层面看到，但不仅仅是这样，甚至在他们做出的一些应用选择上，这些模型公司也变得更加有主见。如果您使用过如Runway或Clean这样的工具，现在您基本可以通过提示来控制相机角度、镜头的宽广度以及其他类似人类电影摄影师会处理的方方面面。您可以提示视频如何在屏幕表面上掠过。这也成为了在同一个视频中可能对不同部分使用不同技术的一个重要因素，这很有趣。

So, I still think IdeaGram is one of the most unique models for sort of what it does, what it's great at, which is text generation, sort of aesthetic that it has. It just sits in a very unique place in the ecosystem. Yeah, we did an internal competition where we had to generate a bunch of video, a 30 second video. And IdeaGram was amazing for that because you could not get that layer of specificity anywhere else. And then you could then take what was generated in IdeaGram and put it into another model to animate it or to swerve or to do whatever you needed to do. Well, they also have a fun feature, which essentially is image to text. So you have a meme or a copyrighted image that you want to replicate or at least be inspired by.
所以，我仍然认为IdeaGram是一个非常独特的模型，在其擅长的文本生成和美学上表现卓越。它在整个生态系统中占据了一个非常独特的位置。我们曾经搞过一个内部比赛，需要生成一段30秒的视频。IdeaGram在这方面表现出色，因为它在细节上的精准度是其他地方无法达到的。你可以将IdeaGram生成的内容放到另一个模型中，进行动画处理或其他需要的操作。此外，它还有一个很有趣的功能，就是“图像转文本”。这意味着如果你有一个模因或是受版权保护的图像，想要复制或至少从中获得灵感，都可以通过这个功能实现。

Yeah. You can use their image to text and then use that text as the prompt to create an image. I also found that fascinating because I would prompt something and as you learn when you're prompting with AI. In general, you learn that you don't know what you're looking for. Yes. And so when I was prompt, IdeaGram would modify your prompt before generating the image. And then you could actually go and interrogate that and be like, oh, that's why I'm getting X, Y, or Z. So video actually in general, it tends to be more of a mobile first phenomena. Right? We see tons of even before AI, tons of applications that focus on creators being able to edit and splice video.
是的，你可以先将他们的图像转成文本，然后再用这个文本作为提示词来生成图像。我觉得这很有趣，因为当你用人工智能进行提示时，你会发现你其实不知道自己在寻找什么。是的。所以当我在用 IdeaGram 提示时，它会在生成图像之前修改你的提示词。然后你可以去查看这些修改，了解为什么会出现某些结果。总的来说，视频通常更偏向于移动端。即使在人工智能出现之前，我们也看到过很多应用专注于让创作者编辑和剪辑视频。

What are we seeing in terms of the difference between what's working on mobile and what's working on desktop? It's somewhat obvious, but like a lot of the things that are working on mobile are either things you want to use on the go or where the underlying asset you're working with is easily captured by the phone. So like all of the avatar apps blew up on mobile because you have 10 selfies of yourself sitting on your phone. A lot of the voice first consumer products that we are seeing working actually are on mobile versus web because it's easier and more natural to talk into your phone for language learning or for companionship or other use cases than it is to maybe talk into your laptop and same with homework helper apps. Like those are really blowing up on mobile as compared to web.
我们在观察手机端和桌面端有何不同之处。这似乎有些显而易见，很多在手机端有效的应用，要么是你在移动中想要使用的，要么是你可以用手机轻松捕捉其基本内容的应用。例如，所有头像类的应用在手机上爆火，因为手机里有十几张自己的自拍。我们看到许多语音优先的消费产品实际上在手机上比在网上更有效，因为在手机上进行语言学习或者陪伴等用途时，使用语音交流更简单自然，而在电脑上可能就不如手机方便。同样，作业助手类的应用在手机上也比在网上流行得多。

So maybe another interesting breakdown that kind of represents where we are in the innovation curve is not just what is getting views but what's actually making money and how those aren't always one to one mirrored. What is making money today where we're learning there and is that the same as what's getting traffic? So for the first time we actually ranked the top 50 by what sensor tower can measure as mobile revenue, which is typically in-app purchases and subscription. So probably not ads. And we ranked those separately from what has the most monthly active users. And there was only 40% overlap between the two lists. So a lot of difference.
也许另一个有趣的分析角度是，我们目前在创新曲线中的位置不只是通过查看哪个内容获得了最多的浏览量，而是分析哪些实际上正在赚钱，而且这两者并不总是一致的。今天赚钱的因素是什么，我们从中学到了什么，这与获得流量的因素是否相同？首次，我们根据Sensor Tower能测量的移动收入（通常是应用内购买和订阅，而不是广告）对前50名进行了排名，并将其与每月活跃用户数量最多的应用分开排名。结果这两份名单之间的重合率只有40%。所以其中有很大的区别。

The surprise to me actually was the main categories are the same in terms of what's making money versus what people are using. So photo and video generators, photo and video editors, beauty filters and beauty enhancer is massive standalone category and then the realm of chat GPT copycat apps both making a ton of money getting a lot of users. But the companies within those categories are very different in terms of who's making money and who has the most usage.
令我感到意外的是，赚钱的主要类别与人们实际使用的类别大致相同。比如，照片和视频生成器、照片和视频编辑器、美颜滤镜和美颜增强器等是非常庞大的独立类别；而类似于ChatGPT的聊天应用也同样赚了很多钱并吸引了大量用户。不过，在这些类别中，各家公司在赚钱能力和用户数量方面却有着很大的差异。

We actually found we plotted like revenue per user versus number of users. And we found the apps that had smaller user bases out of the sample set were much more likely to be making significantly more money on a per user basis. So apps like speak apps like order captions and video editing. There's a lot of reasons for this. One is that if you are making a lot of money per user, you're probably more of a serious prosumer app. And so you've probably actually gated the usage pretty significantly.
我们实际上绘制了一张图表，显示每个用户的收入与用户数量之间的关系。我们发现，在样本集中，用户数量较少的应用程序往往在每个用户的基础上赚取更多的钱。例如，语音应用、字幕制作应用和视频编辑应用。有很多原因导致这种情况，其中一个原因是，如果每个用户为你带来不少收入，你很可能是一个更为专业的专业级消费应用。因此，你可能对应用的使用设置了一定的限制。

Like you have to subscribe to use the product. And so there are companies on here that might be making $50, $100 million in ARR off of only a million users, two million users. So they wouldn't make the ranks ironically enough for monthly active users. But they rank really, really high on a revenue basis, which is exciting. And then as anyone who looks at the mobile list knows there's a lot of maybe for the tech audience seemingly random products on there of like, I've never heard of this. Is this from a startup?
要使用这个产品，你必须订阅。因此，这里有些公司可能只有一百万、两百万用户，但它们的年度经常性收入（ARR）却能达到5000万、1亿美元。因此，仅按月活跃用户数来看，它们可能并不突出。但是如果从收入来看，它们的排名非常高，这点令人兴奋。而且，了解移动应用名单的任何人都知道，名单上有很多产品可能对科技观众来说显得很随机，比如说：“我从没听说过这个。这是某个初创公司的产品吗？”

And on mobile especially, there is a very precise game that you can play with like app store ads and other kind of paid but fairly low-cost acquisition channels. And if you're doing this as an indie developer or maybe an app studio running internationally, you're not looking for the 10x payback of acquisition costs that we might be looking for as venture investors. So if you make back one or two extra money on a user, that's amazing.
在移动端，特别是移动应用商店广告和其他付费但成本较低的获取渠道上，有一种非常精准的策略可以运用。如果你是一个独立开发者，或者是一个在国际上运营的应用工作室，你不需要像风险投资者那样追求获取成本的10倍回报。如果你能在用户身上赚回一到两倍的成本，那就已经很了不起了。

So you can get to 10 million users mostly by paying for them, but you're probably not going to make as much revenue or ultimately as much profit maybe as some of the companies that are lower usage but higher revenue. And is there a learning there in terms of you mentioned how by nature, if you start getting certain features or an application entirely?
所以你可以通过付费的方式获得一千万用户，但相较于那些用户量较低但收入较高的公司，你可能赚不到那么多的收入，最终的利润可能也没那么高。在这方面有没有什么可以学习的？正如你提到的，当开始为某些功能或整个应用增加特性时，会有什么自然的启发吗？

Yeah, you are potentially stifling growth of the overall user base. Is there a learning in terms of how AI founders should be thinking about that trade off today? I think it depends. Some of these markets are naturally maybe not mainstream behavior. Like one example of a category that did appear on the mobile revenue list but was not on mobile usage was several plant identification apps. I love those.
是的，你可能会抑制整体用户群的增长。对于今天的AI创业者来说，他们应该如何思考这种权衡？我认为这取决于市场的性质。有些市场可能天生就不是主流行为。比如有一类应用——植物识别应用，它们出现在移动设备的收入列表上，但并没有出现在使用排行榜上。我非常喜欢这些应用。

Yeah, what you take a picture of you saved on the plan, it tells you exactly what it is. If you've seen that plant before, is that an app that a hundred million people will have on their phones maybe not? But if you're one of the like I can think of a few relatives who love plants or love birds and like totally they'll pay a hundred dollars a year for that and they'll use it every day or every other night.
是的，你拍照后保存下来的东西，它会准确告诉你那是什么。如果你以前见过这种植物，那会是一款拥有上亿用户的应用吗？可能不是。但如果你是那种爱好植物或者鸟类的人，我能想到我有一些亲戚就是这样的人，他们非常乐意每年花一百美元买这个应用，并且每天或隔一天就会使用它。

So I think it's more for founders optimizing for the type of product you have and how mainstream it can be. All right, so there's a lot of information here that we've covered. We've covered desktop, we've covered mobile, we've covered revenue versus users. Yeah. And then we've also talked about the stickiness of some of these players, right?
所以，我认为这更多是为了让创始人优化他们的产品类型以及产品能否进入主流市场。好，我们已经讨论了很多信息。我们谈到了桌面应用、移动应用以及收入和用户之间的对比。是的，我们还讨论了一些平台的用户粘性问题，对吗？

You said there was a 16 that have showed up every single list. So what can we learn from the last few lists? I feel like the biggest thing now being a consumer investor for close to a decade now it's almost like the more you know the less you know in some cases because it all just comes back to the product at the end of the day.
你说有16个人每次榜单都会出现。那么我们能从过去的几个榜单中学到什么呢？我觉得作为一个接近十年的消费者投资者，现在最大的感受是，有时候你知道得越多，反而觉得自己知道得越少，因为到头来，一切都归结于产品本身。

Like technologists or investors can have opinions on the best monetization strategy or the best growth hacks. But in the end, if the product isn't capturing users' attention and isn't retaining them, the business is just going to be completely a leaky bucket of users and users out. Often we meet with these amazing like PhD researchers best in class in the whole world in terms of their technical understanding of a model or capability. And they can struggle building and consumer sometimes because often the more complicated thing is not actually the thing that is highest utility, most delightful, most helpful to a consumer user.
就像科技专家或投资者可以对最佳盈利策略或增长技巧有自己的看法。但最终，如果产品不能吸引用户的注意力，不能留住他们，那么这项业务就会像漏水的桶一样，不断有用户流失。我们经常会见到一些全球顶级的博士研究人员，他们在某个模型或技术方面有非常深的理解。然而，他们有时会在面向消费者的产品开发中遇到困难，因为通常最复杂的东西并不是对消费者用户来说最有用、最令人愉悦、最有帮助的。

So we never like to be prescriptive on consumer products, but in general, we see when teams focus on either the pain point they're trying to solve or the unique experience they're trying to create and build towards that. And if that means your actually the old model is better than the new model, use that if that means it's just one AI feature instead of the whole product being built on AI because it's not stable enough. Like do that. I think in consumer, you really have to let the data view your guide there.
我们通常不喜欢对消费产品做过多的规定，但总体来说，我们发现，当团队专注于解决他们要解决的痛点或是他们希望创造的独特体验时，往往能取得更好的效果。如果说旧模式比新模式更好，那就继续使用旧模式；如果说选择仅在产品中加入一个AI功能会更好，因为基于AI的整体产品尚不够稳定，那就这样做。我认为，在消费产品领域，确实需要让数据来指导你的方向。

Alright, that is all for today. If you did make it this far, first of all, thank you. We put a lot of thought into each of these episodes whether it's guests, the calendar Tetris, the cycles with our amazing editor Tommy until the music is just right. So if you like what we put together, consider dropping us a line at ratethispodcast.com slash a16z. And let us know what your favorite episode is. It'll make my day and I'm sure Tommy's too. We'll catch you on the flip side.
好的，今天就到这里。如果你能坚持看到这里，首先感谢你。我们在每一集的制作上都投入了很多心思，无论是嘉宾的选择、日程安排，还是与我们出色的编辑Tommy一起调整，直到音乐恰到好处。因此，如果你喜欢我们制作的内容，请考虑在ratethispodcast.com/a16z给我们留言，告诉我们你最喜欢哪一集。这将让我们一天都很开心，Tommy也一定会很高兴。我们下次见。