Wix.com Ltd. (WIX) 2026 Q2 法說會逐字稿

Good morning and thank you for standing by. Welcome to the Wix second quarter conference call. (Operator Instructions) Please be advised that today's conference call is being recorded.

早安，感謝各位撥冗等候。歡迎參加 Wix 第二季電話會議。(接線員指示) 請注意，今天的電話會議將被錄音。

I would now like to hand the conference over to your first speaker today, Emily Liu, Investor Relations. Please go ahead.

現在我想把會議交給今天的第一位講者，投資人關係部的 Emily Liu。請開始。

Thanks, and good morning, everyone. Welcome to Wix's second-quarter 2026 earnings call. Joining me today to discuss our results are Avishai Abrahami, CEO and Co-founder; Nir Zohar, President and Co-Founder; and Lior Shemesh, our CFO.

謝謝，各位早安。歡迎參加 Wix 2026 年第二季財報電話會議。今天與我一同討論業績的有：共同創辦人暨執行長 Avishai Abrahami、共同創辦人暨總裁 Nir Zohar，以及財務長 Lior Shemesh。

During this call, we may make forward-looking statements, and these statements are based on current expectations and assumptions. Please consider the risk factors included in our press release and most recent Form 20-F that could cause our actual results to differ materially from these forward-looking statements.

在本次電話會議中，我們可能會發表前瞻性陳述，而這些陳述係基於目前的預期與假設。請參考我們新聞稿及最新 Form 20-F 中所列的風險因素，這些因素可能導致實際結果與前瞻性陳述出現重大差異。

We do not undertake any obligation to update these forward-looking statements except as required by law. In addition, we will comment on non-GAAP financial results and key operating metrics. You can find all reconciliations between our GAAP and non-GAAP results in the earnings materials and in our Interactive Analyst Center on the Investor Relations section of our website, investors.wix.com.

除非法律要求，否則我們不承擔更新這些前瞻性陳述的任何義務。此外，我們將評論非 GAAP 財務結果與關鍵營運指標。您可在財報資料以及我們網站投資人關係專區 investors.wix.com 的互動式分析師中心，查閱 GAAP 與非 GAAP 結果之間的所有調節表。

With that, I'll turn the call over to Avishai.

接下來，我把電話交給 Avishai。

Thanks, Emily. As I mentioned in our shareholder letter, our strategy is centered on deepening our ownership of the technology stack behind our AI products while expanding Wix's role across the AI ecosystem and throughout the entire user journey from ideation online to fully operational businesses and deployed software. We're pursuing this through two distinct purpose-built paths rather than betting on a single outcome for how AI reshapes online creation.

謝謝，Emily。如我在致股東信中提到的，我們的策略核心在於加深我們對 AI 產品背後技術堆疊的掌控，同時擴大 Wix 在 AI 生態系中的角色，並覆蓋從線上構想到完整營運的企業與已部署軟體的整個使用者旅程。我們透過兩條各自為特定目的打造的路徑並行推進，而不是押注 AI 將如何重塑線上創作的單一結果。

Wix Harmony continues to serve self-creators through visual drag-and-drop creation paired with AI. Base44 serves a different and fast-growing motion, natural language, vibe-coded app and software creation. Running both in parallel means we capture demand wherever the market evolves. We achieved a significant product milestone this quarter in support of this strategic priority with the introduction of Base 1, the first platform in its category to launch a proprietary LLM.

Wix Harmony 持續以視覺化拖放式創作結合 AI，服務自助創作者。Base44 則服務另一股快速成長的趨勢：以自然語言、vibe-coded 的方式進行應用程式與軟體創作。兩者並行運作，意味著無論市場如何演進，我們都能掌握需求。本季我們在支持此策略優先事項上達成重要產品里程碑，推出 Base 1——該類別中首個推出自有專有 LLM 的平台。

Building this model in-house was a deliberate choice. It gives us tighter control over quality, faster iteration cycles and a better long-term cost structure. It also means every interaction on the platform continues to compound into a proprietary data asset, one that keeps improving Base 1 rather than benefiting an external vendor's model.

選擇在內部打造這個模型是我們刻意的決策。這讓我們能更緊密地掌控品質、更快的迭代週期，以及更佳的長期成本結構。同時也代表平台上的每一次互動都會持續累積成為專有資料資產，讓 Base 1 不斷改進，而不是讓外部供應商的模型受益。

On the economics, Base 1 is also driving tangible margin improvement for Base44. As we shift more inference to our own model and continue cost optimization work, we see AI-related costs fall across free and paid users, improving the Base44 gross margin from roughly 0% when we started the year to approximately 60% in the second-half of the year.

在經濟效益方面，Base 1 也正為 Base44 帶來實質的毛利率改善。隨著我們將更多推論轉移到自有模型並持續進行成本最佳化，我們看到免費與付費使用者的 AI 相關成本同步下降，使 Base44 毛利率從年初剛開始時約 0%，提升至下半年約 60%。

This trajectory reflects the business moving past early-stage hyper-growth economics toward a more sustainable, margin-healthy model with further room to improve. Base 1 also demonstrates what is possible when bringing together Wix's 20 years of expertise building and making complex technology accessible to everyone with Base44's nimble and ambitious AI-powered workflow.

這一走勢反映出業務正從早期超高速成長階段的經濟模型，邁向更可持續、毛利更健康的模式，且仍有進一步改善空間。Base 1 也展示了當 Wix 20 年來在建置並讓複雜技術普及化的專業，與 Base44 靈活且具企圖心的 AI 驅動工作流程結合時，所能達成的可能性。

The same data science organization that led the development of the Wix Harmony LLM, our first purpose-built AI model introduced earlier this year was able to support the Base44 team directly with existing research, infrastructure and experimentation capabilities achieved over years of AI development.

同一個主導 Wix Harmony LLM 開發的資料科學組織——我們今年稍早推出的第一個為特定目的打造的 AI 模型——能夠直接以多年 AI 研發累積的研究成果、基礎設施與實驗能力，支援 Base44 團隊。

This team, which I work closely with, played an invaluable role in the creation of Base 1 and was a major reason we were able to bring our model into production so quickly. With continued demand strength and now more control over the technology stack powering the platform, we are doubling down on a strategy that has already proven successful.

我與這個團隊密切合作；他們在 Base 1 的打造過程中扮演了無可取代的角色，也是我們能如此快速將模型導入量產的主要原因之一。在需求持續強勁、且我們如今對驅動平台的技術堆疊擁有更高掌控度之下，我們正加碼投入一項已被證明成功的策略。

The vision behind Wix Harmony is focused on building AI and human collaboration into the core editing experience helping users generate, refine and iterate on content while staying fully in control of the final result.

Wix Harmony 背後的願景，是將 AI 與人類協作融入核心編輯體驗，協助使用者生成、精修並反覆迭代內容，同時仍能完全掌控最終成果。

This reflects the broader principle behind everything we build. As the market, the technology and user expectations keep shifting, our focus is to continuously evolve our platform to the needs of our users. We've continued to make significant advancements in Wix Harmony as part of this ongoing commitment. Our focus now is execution, leaning into our position of strength to create products that generate real value for users and durable growth for Wix.

這也反映了我們打造一切產品背後更廣泛的原則。隨著市場、技術與使用者期待持續變動，我們的重點是持續演進平台以符合使用者需求。作為這項長期承諾的一部分，我們在 Wix Harmony 上持續取得重大進展。我們現在的重點是執行，善用我們的優勢地位，打造能為使用者創造真實價值、並為 Wix 帶來持久成長的產品。

With that, I'll turn it over to Nir.

接下來，我把電話交給 Nir。

Thanks, Avishai. I want to start with a quick overview of the growth trends we're seeing across the business, then explain how those insights inform our investment strategy. We continue to execute on our plan in the second quarter with steady top-line performance.

謝謝，Avishai。我想先快速概述我們在整體業務中看到的成長趨勢，接著說明這些洞察如何影響我們的投資策略。第二季我們持續按計畫執行，營收端表現穩健。

Bookings grew 12% year-over-year and revenue grew 15% year-over-year, driven by strong Base44 performance and continued growth in our core Wix business in line with our expectations with encouraging performance from Wix Harmony.

預訂額（Bookings）年增 12%，營收年增 15%，主要由 Base44 的強勁表現帶動，同時我們核心 Wix 業務也依照預期持續成長，Wix Harmony 的表現亦令人鼓舞。

As a result, year-over-year self-creators revenue growth accelerated sequentially to 14% this quarter, underpinned by healthy business fundamentals, including improving conversion from free-to-paid users, stable retention behavior, and a robust top-of-funnel. Partners revenue grew 17% year-over-year, in line with the expectations we provided in early June.

因此，本季自助創作者（self-creators）營收年增率較前一季加速至 14%，其背後有健康的業務基本面支撐，包括免費轉付費的轉換率改善、留存行為穩定，以及強勁的漏斗頂端（top-of-funnel）。合作夥伴（Partners）營收年增 17%，符合我們在 6 月初提供的預期。

Encouragingly, we saw a step-up in Base44 contribution to the partner segment in the second quarter as professionals increasingly leveraged AI and AI agents in their workflows and expanded their pipelines to include software creation.

令人振奮的是，第二季我們看到 Base44 對合作夥伴分部的貢獻提升，因為專業人士在工作流程中愈來愈多地運用 AI 與 AI 代理（AI agents），並將其案源管線擴展至包含軟體創作。

GPV in Q2 grew 3% year-over-year, driven primarily by the wind-down of our subsidiary, InkFrog. As part of our organizational realignment in June to refocus efforts on high-return products, InkFrog had moderate contribution to GPV but low monetization. Its wind down improves our mix towards better monetized GPV dollars, resulting in better take rate.

第二季 GPV 年增 3%，主要受到我們子公司 InkFrog 逐步收尾的影響。作為我們 6 月組織調整、重新聚焦高報酬產品的一部分，InkFrog 對 GPV 有中度貢獻，但變現能力偏低。其收尾將改善我們的組合，轉向變現更佳的 GPV 金額，從而帶來更好的抽成率（take rate）。

After our June outlook adjustments related to our partners business, I want to make it clear that our partners business remains a key area of investment and focus. We are already actively testing new solutions with our agency partners and believe that these solutions will align our platform with how the partners and agencies' ecosystem is evolving, turning to Base44, which continued on its strong growth trajectory in Q2.

在我們 6 月針對合作夥伴業務調整展望之後，我想明確表示：合作夥伴業務仍是我們投資與聚焦的關鍵領域。我們已在與代理商合作夥伴積極測試新解決方案，並相信這些方案將使我們的平台與合作夥伴及代理商生態系的演進方向一致；同時，Base44 在第二季也延續其強勁的成長軌跡。

Top of final demand remained elevated with the newest cohort outperforming the previous one, while renewal activity led overall Base44 growth. Encouragingly, we continue to see more new and existing users choose annual plans as they increasingly trust Base44 for their software needs.

最終需求的漏斗頂端（top of final demand）仍維持在高檔，最新一批用戶群的表現優於前一批，而續約活動帶動了 Base44 的整體成長。令人鼓舞的是，隨著使用者愈來愈信任 Base44 來滿足其軟體需求，我們持續看到更多新舊用戶選擇年繳方案。

Given the continued strong demand, meaningful product improvements and the significantly improved margin profile of the Base44 business driven by Base 1, which Lior will speak to in more detail shortly, we plan to invest further into Base44.

鑑於需求持續強勁、產品顯著改善，以及由 Base 1 驅動、Base44 業務毛利結構大幅提升（Lior 稍後會更詳細說明），我們計畫進一步加大對 Base44 的投資。

As a result, we are raising our TROI target moderately, underscoring our confidence that the opportunity in front of us remains competitive but massive. Our priority is to aggressively capture Base44 market share as the AI-powered app creation space continues to be dynamic and growing, with strong belief that the strategic investments we're making today will drive sustained growth and market leadership over the long-term.

因此，我們將適度上調 TROI 目標，凸顯我們對眼前機會的信心：競爭激烈但規模龐大。我們的優先事項是積極搶占 Base44 市占率，因為 AI 驅動的應用程式創作領域仍然充滿動態且持續成長；我們堅信，今天所做的策略性投資將在長期推動持續成長與市場領導地位。

With that, I'll hand it over to Lior, who will discuss how we expect these priorities to flow through the financials. Lior?

說到這裡，我把時間交給 Lior，他將說明我們預期這些優先事項將如何反映在財務表現上。Lior？

In the second quarter, we delivered continued solid top-line growth and continue to position the business for long-term success and free cash flow generation by innovating our platform and managing the business with discipline. You just heard from Nir about our strategic priorities and top-line trends, so I'll focus my remarks on the cost side of the business, where we drove strong execution on initiatives that I believe will create meaningful leverage over time.

在第二季，我們持續交出穩健的營收成長，並透過平台創新與紀律性的營運管理，持續為企業的長期成功與自由現金流的產生做好布局。各位剛剛已聽 Nir 談到我們的策略優先事項與營收趨勢，因此我將把重點放在成本端；在這裡，我們在多項計畫上展現強勁的執行力，我相信這些將隨時間帶來顯著的營運槓桿。

I am proud to say that we have delivered on a key initiative planned for this year, a company-wide priority for 2026 was to lower inference costs with our own LLMs. We started with our Harmony model earlier this year. Then in June, we launched our Base 1 model, which we believe structurally improves the margin profile of the business, and we are seeing immediate results.

我很自豪地說，我們已完成今年規劃的一項關鍵計畫；公司層級對 2026 年的優先事項之一，就是透過自有 LLM 降低推理成本。我們在今年稍早先從 Harmony 模型開始。接著在 6 月，我們推出了 Base 1 模型；我們相信這在結構上改善了公司的毛利結構，而且我們已看到立竿見影的成果。

We now expect non-GAAP gross margin for Base44 to be approximately 60% in the second-half of this year. This is a very significant improvement from the near-zero gross margin entering 2026. These cost savings are expected to translate into approximately 2 points of total non-GAAP gross margin improvement in the second-half of the year versus the first half for the consolidated business.

我們目前預期，今年下半年 Base44 的非 GAAP 毛利率約為 60%。相較於進入 2026 年時接近零的毛利率，這是非常顯著的改善。預期這些成本節省將轉化為：相較於上半年，合併口徑的下半年總非 GAAP 毛利率約提升 2 個百分點。

Turning quickly to the second quarter, let's start with gross margin. Our second quarter total non-GAAP gross margin was 67%, a slight increase sequentially and down 3 points year-over-year. Our lower year-over-year total non-GAAP gross margin was driven by continued elevated investments in Base44 to support its rapid growth and elevated AI compute costs as we scale and maximize gross profit dollars.

快速回到第二季，先從毛利率談起。我們第二季總非 GAAP 毛利率為 67%，較前一季小幅上升，但年減 3 個百分點。年對年總非 GAAP 毛利率下降，主要是因為我們持續對 Base44 進行較高投入以支援其快速成長，以及在我們擴張規模並最大化毛利金額（gross profit dollars）的同時，AI 運算成本仍然偏高。

Our second quarter margin reflects stable gross margins in our core Wix business compared to the prior year period. Total non-GAAP operating income came in at 12% of revenue, primarily driven by continued higher levels of sales and marketing expenses in the quarter. Non-GAAP S&M expenses remained elevated in the second quarter as expected as we continue to accelerate marketing investments into Base44 in order to capture strong top-of-funnel demand trends throughout the quarter, while also seeing AI inference and compute costs associated with free Base44 users continue to increase sequentially.

我們第二季的毛利率也反映出：相較去年同期，我們核心 Wix 業務的毛利率維持穩定。總非 GAAP 營業利益為營收的 12%，主要受到本季銷售與行銷費用持續較高所影響。如預期，第二季非 GAAP 銷售與行銷（S&M）費用仍維持高檔，因為我們持續加速對 Base44 的行銷投資，以掌握本季整體漏斗上端（top-of-funnel）強勁的需求趨勢；同時，與免費 Base44 使用者相關的 AI 推理與運算成本也持續較前一季增加。

As Nir discussed, with the margin profile of Base44 fundamentally improved, we are going to lean into S&M expenses to capture increased demand. As a result, we plan to reinvest the entirety of the gross margin savings I just discussed back into sales and marketing activities.

如 Nir 所提到，隨著 Base44 的毛利結構已從根本上改善，我們將加大銷售與行銷費用的投入，以掌握更高的需求。因此，我們計畫把我剛才提到的毛利節省金額全數再投資回銷售與行銷活動。

As a result, we anticipate third and fourth quarter sales and marketing activities to remain elevated as we now aim toward a moderately longer TROI, finally turning to our balance sheet. We ended Q2 26 with approximately $960 million in cash and cash equivalents and $1.63 billion in short and long-term debt.

因此，我們預期第三季與第四季的銷售與行銷活動將維持在較高水準，因為我們現在的目標是略為拉長的 TROI（投資回收期）。最後談到資產負債表。我們在 26 年第二季末持有約 9.6 億美元的現金及約當現金，以及 16.3 億美元的短期與長期負債。

Let's turn now to the outlook for the third quarter and second-half of 2026. We are maintaining our current guidance and continue to expect bookings to grow at a low 10s percentage, lagging revenue growth by a few points, and revenue to grow at a low to mid-teens percentage on a year-over-year basis for the full year.

接著談第三季與 2026 年下半年的展望。我們維持目前的財測指引，並持續預期預訂（bookings）將以 10% 出頭的百分比成長，較營收成長落後幾個百分點；同時，全年營收將以年對年低到中段十位數百分比成長。

We expect third quarter revenue to grow at the low double-digit percentage on a year-over-year basis. For the full year 2026, we are maintaining our expectation for free cash flow margin, excluding acquisition and restructuring costs to be in the high 10s.

我們預期第三季營收將以年對年低雙位數百分比成長。就 2026 全年而言，我們維持對自由現金流利潤率的預期（不含併購與重組成本）為 10% 高段。

Our full-year outlook assumes that the approximately 2 points of non-GAAP gross margin improvement in the second-half in the consolidated business will be reinvested into Base44 sales and marketing through the rest of the year. This reflects our expectation that demand for Base44 will remain elevated, enabling us to capture additional market share as the business continues to outperform.

我們的全年展望假設：合併口徑下半年約 2 個百分點的非 GAAP 毛利率改善，將在今年剩餘期間再投資於 Base44 的銷售與行銷。這反映我們預期 Base44 的需求將維持高檔，使我們能在業務持續優於預期的同時，取得更多市占。

We expect to offset this increased sales and marketing investment in Base 44 with lower AI costs and decreased marketing costs for core Wix in the second-half of the year, in line with seasonality and lapping the Super Bowl investments earlier this year. We expect R&D expenses to remain stable in the second-half of the year compared to the first half as the FX headwind from a strengthening Israeli shekel offset savings from our organizational realignment. As a result, we continue to expect non-GAAP operating margin for the consolidated company to step up in the second-half of the year, putting us on track to achieve our free cash flow outlook.

我們預期將以 Base44 較低的 AI 成本，以及核心 Wix 在下半年較低的行銷成本來抵銷這項增加的 Base44 銷售與行銷投資；這符合季節性因素，且今年稍早的超級盃（Super Bowl）投資也將進入比較基期（lapping）。我們預期研發（R&D）費用在下半年相較上半年將維持穩定，因以色列謝克爾走強帶來的匯率逆風，將抵銷我們組織調整所帶來的節省。因此，我們仍預期合併公司的非 GAAP 營業利益率將在下半年提升，使我們有望達成自由現金流展望。

In conclusion, our conviction in our near-term strategy and ambitious AI product roadmap remains unchanged. The team is incredibly focused on executing our shared vision and I am confident that key initiatives like Harmony and Base44 are the right areas of investment. We are utilizing this year to lean into our future growth and leverage AI across every function to drive higher output.

總結來說，我們對近期策略與具雄心的 AI 產品藍圖的信心依然不變。團隊高度專注於執行共同願景，我也有信心 Harmony 與 Base44 等關鍵計畫是正確的投資方向。我們正利用今年加大對未來成長的投入，並在各個職能全面運用 AI，以提升產出。

The decisions we are making today will pay off in the long-term as we continue to build, refine and deliver products that capture additional market share and drive compounding financial performance for our shareholders.

我們今天所做的決策，將在長期帶來回報；我們會持續打造、精進並交付能取得更多市占的產品，並為股東帶來複利式的財務表現。

Operator, we are now ready for questions.

接線員，我們現在準備開始問答。

(Operator Instructions)

（接線員指示）

Ygal Arounian, Wedbush.

Ygal Arounian，Wedbush。

Hey, good morning guys. I want to first dig in on Base 1 and the proprietary LLM and maybe if you could expand on the product in general, how it's built, maybe if there's any quality trade-offs and potential for further AI cost improvements, compute cost improvements over time. Just love to get a little bit more detail on that and then have a follow-up.

嗨，大家早安。我想先深入問一下 Base 1 與自研 LLM；也許你們可以更全面地說明這個產品：它是如何建構的、是否存在任何品質取捨，以及未來在 AI 成本與運算成本方面是否還有進一步改善的空間。很希望能聽到更多細節，之後我還有一個追問。

Of course. So Base 1 is a model that we train in-house. It is a big part of the concept is that you will always keep training it based on what we see that works better for our users. And this is a lot where the secret choice is. How do you know what is good? How do you know what is bad? Because essentially every time you write a prompt (inaudible) Base44, the model usually generates thousands or hundreds of thousands of source code lines and we have to somehow know how to pick the good results as the bad results.

當然可以。Base 1 是我們在內部訓練的模型。這個概念很重要的一部分是：我們會根據觀察到哪些對使用者更有效，持續對它進行訓練。而這也正是關鍵訣竅所在。你要如何知道什麼是好的？你要如何知道什麼是不好的？因為本質上，每次你在 Base44 寫一個提示詞（聽不清），模型通常會生成數千行或數十萬行的原始碼，而我們必須想辦法知道如何從中挑出好的結果、淘汰不好的結果。

So once you do that, we're able to push that back into the model and keep training it to continuously be better. Today, maybe because of this approach, we're in a place that we can already see that the result of Base 1 are better than any of the models that we have from frontier providers. It's not better for everything. I want to be very clear about it. It's better for Base44, right? But we do see a significant improvement.

一旦做到這點，我們就能把這些回饋再送回模型，持續訓練它，讓它不斷變得更好。今天，或許正因為這種方法，我們已經處在一個位置：我們可以看到 Base 1 的結果優於我們從前沿供應商（frontier providers）取得的任何模型。但它不是在所有方面都更好。我想非常清楚地說明這點。它是對 Base44 更好，對吧？但我們確實看到顯著的提升。

And of course, the cherry on the top is the fact that it also costs dramatically less. And we think that this in, it's part of a very long-term strategy where we'll be able to continue and improve Base 1 as a result of that, continuously improve one of the critical parts of Base44, making Base44 a better product.

當然，錦上添花的是，它的成本也大幅更低。我們認為這是非常長期策略的一部分；藉由這樣的方式，我們將能持續改進 Base 1，並因此持續強化 Base44 的關鍵組成之一，讓 Base44 成為更好的產品。

Okay, great. Thanks. And can you maybe just give an update also on what you're seeing in the partner ecosystem as more builds are going towards five coding and AI builds and how's the competitive environment changing around that as well, maybe with Base44 in particular, but just more broadly and how, those that cohort is building and starting sites and online experiences. Thank you.

好的，很棒。謝謝。另外，你能否也更新一下你們在合作夥伴生態系中看到的情況：隨著越來越多的建置轉向「氛圍編碼（vibe coding）」與 AI 建置，競爭環境也在如何改變？也許特別針對 Base44，但也更廣泛地談談；以及那個族群如何在建置並開始建立網站與線上體驗。謝謝。

Hey, it's, Nir. So I think I'm not going to go back to what we already shared, but I do want to talk more broadly on kind of the dynamics we're seeing. There's definitely more and more appetite also for the part on the partners ecosystem for AI solutions and AI products.

嗨，我是 Nir。所以我想我不會再回到我們已經分享過的內容，但我確實想更宏觀地談談我們看到的一些動態。很明顯地，對於合作夥伴生態系中 AI 解決方案與 AI 產品的那一部分，需求與意願也越來越高。

We're seeing this across the Board, we're seeing this obviously with the ones who are using Harmony on the weak side, we're seeing this with more and more than using Base44, some of which to create websites, some of which to create and build applications for their clients.

我們在各方面都看到了這點；很明顯地，在較弱勢端使用 Harmony 的那些客戶身上也看得到；也看到越來越多人在使用 Base44，有些用來建立網站，有些用來為他們的客戶建立與開發應用程式。

And we're also definitely seeing it in those kind of conversations. We're doing with all of our kind of advanced beta testers on new products that we intend to release to the hands of the partners. So I think from that standpoint, we're probably going to keep on seeing that. And our goal is to be able to deliver value across the full portfolio of our products and offerings.

而且我們在這類對話中也確實看到了。我們正與所有進階的 Beta 測試者就我們打算釋出、交到合作夥伴手上的新產品進行交流。所以我想從這個角度來看，我們可能會持續看到這樣的趨勢。而我們的目標是能在我們完整的產品與服務組合上交付價值。

Thank you, guys.

謝謝各位。

Elizabeth Porter. Elizabeth.

Elizabeth Porter。Elizabeth。

Great. Thank you so much for the question. I wanted to follow-up again on the Base44 gross margin improvement, which is really impressive. So how much of the expected improvement to the approximately 60% non-GAAP gross margin in the back half of the year has already begun to be reserved versus remaining kind of in the forecast period? And what percentage of inference traffic is currently running through Base 1? Are there any constraints we should consider before broader deployment?

很好。非常感謝這個問題。我想再追問一下 Base44 毛利率的改善，這真的很令人印象深刻。所以，預期今年下半年非 GAAP 毛利率約 60% 的改善之中，有多少已經開始反映（或已經鎖定）了，還有多少仍屬於預測期間內尚待實現的部分？另外，目前有多少比例的推論（inference）流量是透過 Base 1 在跑？在更廣泛部署之前，有沒有任何我們需要考量的限制？

Hey, Elizabeth, it's Lior. What I can tell you that the range that we provided for the gross margin, we already see it. I think that this is why we feel a lot of confidence about what we've managed to do. And it was not happening like in day one. It happened gradually over the last few weeks. But we've already started to see the increase in gross margin within the range that I already provided within my guidance.

嗨，Elizabeth，我是 Lior。我可以告訴你的是，我們提供的毛利率區間，我們已經看到了。我想這也是為什麼我們對目前做到的成果很有信心。而且這不是第一天就發生的。它是在過去幾週逐步發生的。但我們已經開始看到毛利率提升，落在我在指引中提供的區間之內。

And then just as a follow-up, I believe Base44's earliest annual cohorts are approaching or starting to pass their first renewal period. So, just curious what you could say about the renewal behavior relative to some of the monthly cohorts or core Wixs, and what are some of the cohort indicators that give you confidence that the Base44 ARR is durable? Thank you.

再追問一下，我相信 Base44 最早的一批年度 cohort 正接近或開始度過他們的第一次續約期。所以想請教，你們能否談談續約行為相較於某些月度 cohort 或 Wix 核心產品的表現如何？以及有哪些 cohort 指標讓你們有信心 Base44 的 ARR 具有持久性？謝謝。

Hey, Elizabeth. So, yeah, you're correct. We are nearing kind of the first round of the annual renewals for Base44 and naturally we're not there yet, so we are not ready to comment or speak about it.

嗨，Elizabeth。是的，你說得沒錯。我們正接近 Base44 第一輪年度續約期，當然我們還沒真正走到那一步，所以目前還不準備對此發表評論或說明。

In terms of what we're seeing in terms of the monthly behavior on Base44, we commented on this in the past. Naturally, it is not at the same rate as Wix, which makes a lot of sense. Wix is a brand and a product that has been existed for many years.

就我們在 Base44 看到的月度行為而言，我們過去也評論過。自然地，它不會和 Wix 是同樣的水準，這很合理。Wix 是一個已經存在多年的品牌與產品。

We have a much higher annual rate there, people know the brand very well. That being said, we are seeing an improvement almost every month in terms of the behavior there and we take all of that into consideration when we are assuming our growth trajectory both for the ARR as well as how we spend our TROI. So we have, I would say, we feel comfortable.

在那邊我們的年度比例高得多，大家也非常熟悉這個品牌。即便如此，我們確實看到那邊的行為幾乎每個月都在改善；而我們在假設 ARR 的成長軌跡，以及我們如何投入 TROI（總投資報酬）時，都會把這些因素納入考量。所以我會說，我們感到放心。

Great, thank you.

很好，謝謝。

Alexei Gogolev, JPMorgan.

Alexei Gogolev，摩根大通（JPMorgan）。

Thank you. Hello, everyone. Can you hear me?

謝謝。各位好。聽得到我嗎？

Yes.

聽得到。

Great. Thank you. I wanted to follow-up on Elizabeth's question about traffic running through Base 1. Any update there? And generally, like what are the biggest drivers of Base44 demand right now, maybe you can talk about user types or channels and what indicators tell you that demand remains elevated?

很好。謝謝。我想延續 Elizabeth 關於透過 Base 1 跑的流量的問題。這方面有任何更新嗎？另外更廣泛地說，目前推動 Base44 需求的最大驅動因素是什麼？也許你可以談談使用者類型或渠道，以及哪些指標讓你們判斷需求仍然維持在高檔？

We already are running significant traffic on models that are not the classic frontier models. But saying that, we continuously now test what is the right balance and how to effectively use each model that we have. It's very complicated because we were always trying to estimate what is the best model to solve that specific application that the user is trying to build, right? And that varies a lot.

我們已經在非傳統的前沿（frontier）模型上跑了相當可觀的流量。不過即便如此，我們現在仍持續測試什麼才是正確的平衡，以及如何有效地使用我們擁有的每一種模型。這非常複雜，因為我們一直在嘗試估算：要解決使用者正在建構的特定應用，哪一個模型是最佳選擇，對吧？而這差異很大。

But I think that the interesting part is that our ability to predict better which models to use is also improving. The biggest drivers for Base44 demand is actually the satisfaction of users from the application that they built because what we see is that when somebody is successfully building an application and is happy with what he built, he'll share it with his friends, he'll share it with coworkers, and that is the biggest driver of new users, in addition, of course, to marketing.

但我認為有趣的是，我們更準確預測該用哪些模型的能力也在提升。Base44 需求最大的驅動因素，其實是使用者對他們所建構的應用程式的滿意度；因為我們看到，當某個人成功建出一個應用、並且對成果感到滿意時，他會分享給朋友、分享給同事，而這就是新使用者最大的來源，當然再加上行銷。

Because of that, we are such big believers in the product quality that the better the product quality is, the better the demand will become. And we've seen that in the early days of Wix, of course. So we are very familiar with this pattern of behavior.

也因此，我們非常相信產品品質：產品品質越好，需求就會變得越強。而我們在 Wix 的早期當然也看過這種情況。所以我們對這種行為模式非常熟悉。

In terms of user times, well, I would say about 40% of our users are trying to build personal projects and things that are related to their life or their personal goals in life and 60% are business oriented applications, people taking that what they do at work.

就使用者類型而言，我會說大約 40% 的使用者是在做個人專案，以及與他們生活或個人目標相關的事情；另外 60% 是以商業導向的應用為主，也就是人們把他們在工作上做的事情帶進來。

In addition, when you look at the business segment, you see a big variety. However, I want to maybe one thing to worth emphasizing is that we have more enterprises than we have at Wix, so we ended up having more large businesses on Base44 than we have at Wix.

此外，當你看商業區隔時，會看到非常多樣化。不過我想有一點值得強調的是：我們在 Base44 上的企業客戶比 Wix 還多，所以我們在 Base44 上的大型企業數量比 Wix 還多。

Beyond that, it's a big mix and I think the fact that it's a big mix is a huge part of the strength of the product, just showing that it's versatile enough that it can be used by many kinds of people in many different scenarios, and that is a big way for us to appreciate the potential strength of Base44.

除此之外，整體是很大的混合；而我認為這種「混合度很高」正是產品強項的重要一部分，顯示它足夠多用途，能被各種人在各種不同情境中使用；這也是我們評估 Base44 潛在實力的一個重要方式。

Thank you very much, Avishai. And just a quick follow-up, to what extent is partner activity migrating towards Base44 and how do you ensure that this increases total lifetime value rather than shifting revenue between those buckets?

非常感謝你，Avishai。再快速追問一下：合作夥伴活動在多大程度上正轉向 Base44？以及你們如何確保這會提升總終身價值（LTV），而不是只是在不同收入桶之間移轉收入？

What we see now is a beginning, and it might be a big trend, but we don't know yet, right? It's very early. It's kind of hard to estimate how the world will evolve with different technologies, AI technologies, because it never happened before. What we can say is that we do see some partners that reduce their activity on Wix, and then a lot of them have increased their activity on Base44.

我們現在看到的是一個開始，這可能會成為一個大趨勢，但我們還不確定，對吧？現在還非常早。要估計世界會如何隨著不同技術、AI 技術而演進，其實很難，因為以前從未發生過。我們能說的是，我們確實看到一些合作夥伴降低了他們在 Wix 上的活動，而其中很多人增加了他們在 Base44 上的活動。

I want to be clear, it's not that people are just moving between Base44 to Wix and Wix to Base44, right? We see that also because they use some of our competitors and agencies there are moving to Base44 and I'm sure some of our partners, right, have left using different products in the vibe coding offering. But we do see this migration. I think it will be very irresponsible for me now to try and predict how it evolves in the next year, but we update and when we have more information.

我想說清楚：並不是人們只是在 Base44 與 Wix 之間互相移動，對吧？我們也看到這種情況，因為他們也會使用我們的一些競爭對手，而且有些代理商正在轉向 Base44；我也確定我們的一些合作夥伴，對吧，也離開去使用其他在 vibe coding 產品上的不同方案。但我們確實看到這種遷移。我認為我現在若試圖預測未來一年它會如何演變，是非常不負責任的；但我們會在有更多資訊時再更新。

Thank you very much, Avishai.

非常感謝你，Avishai。

[Stefano Chris, Adidami Company].

[Stefano Chris，Adidami 公司]。

Hi. Just wanted to follow-up on that last question. As you see some partners shifting to Base44, can you talk about the difference in economics for selling the switches, maybe just on revenue or margin contribution? Thank you.

嗨。我只是想跟進上一個問題。當你看到一些合作夥伴轉向 Base44 時，你能談談在銷售這些切換（switches）時經濟性上的差異嗎？也許就收入或毛利貢獻來說。謝謝。

It's very, I think this is the answer. There's no clear answer. It depends what they're doing. A lot of those partners actually ended up doing more heavy applications for the customers. In fact, in my conversation with partners, one of the things they were very happy about is that from building a website and charging $1,000, now they can build application and charge $15,000 or $30,000, okay?

這個問題非常……我想這就是答案。沒有一個明確的答案。這取決於他們在做什麼。很多這些合作夥伴其實最後為客戶做了更重度的應用。事實上，在我和合作夥伴的對話中，他們很開心的一點是：以前做一個網站收 1,000 美元，現在他們可以做應用程式，收 15,000 或 30,000 美元，對吧？

And so I think, again, even for them, a lot of it is new and it's being reshaped. So, but in terms of the general value and the economy aside for weeks, there is that I would say at this stage I would pretty (inaudible) because you might build less projects but more expensive projects and you utilize more of Base44, then the result of that is that architect from that is a bit bigger. However, this is yet to be proven and we need to wait a few more at least months to be able to come back with a clear number.

所以我想，再次強調，即使對他們而言，很多事情也是新的，正在被重新塑造。不過就整體價值與經濟性而言（先不談幾週內的波動），我會說在這個階段我會相當（聽不清），因為你可能做的專案數量變少，但專案單價更高，而且你會更多地使用 Base44；其結果是，從架構角度來看，那部分會更大一些。然而，這點仍有待驗證，我們至少還需要再等幾個月，才能回來給出一個明確的數字。

With regard to the margins, I think that right now when we started to use Base 1. I think that right now Base44 is more or less the same profitability as many other software companies. I think that also there is a lot of there is some I believe that there is more room for improvement. So definitely the gross margin is and profitability is much better than what it was like even just a few months ago.

關於毛利率，我認為目前當我們開始使用 Base 1 時……我認為目前 Base44 的獲利能力大致上與許多其他軟體公司相當。我也認為還有很多……我相信仍有一些、而且有更多改善空間。所以，整體而言，毛利率與獲利能力確實比起甚至就在幾個月前的狀況要好得多。

Got it. Thank you. And just to follow-up on the Base 1 cost savings, is that expected to only help Base44 margins or can that help the rest of the company? Thank you.

了解。謝謝。另外再追問一下，關於 Base 1 的成本節省，這預期只會幫助 Base44 的毛利率，還是也能幫助公司其他部分？謝謝。

Lior, go ahead.

Lior，你來。

Base 1 used for Base44, we did the same thing, if you guys remember, with Harmony. And with Harmony, we also use our own model that we managed to reduce costs significantly. So you can say that we're using the same strategy across the board, and this is why we saw increase in profitability in the second-half of the year compared to the first half.

Base 1 是用在 Base44 上；如果各位還記得，我們在 Harmony 上也做了同樣的事。而在 Harmony 上，我們也使用了我們自己的模型，成功大幅降低成本。所以你可以說我們在各個面向都採用相同策略，這也是為什麼我們看到下半年相較上半年獲利能力有所提升。

Great, thank you.

很好，謝謝。

Josh Beck, Raymond James.

Josh Beck，Raymond James。

Yes, thank you for taking the question. I wanted to ask about the TROI framework. I believe historically it's been less than 12 months, and with the opportunity you see ahead, you're kind of willing to lean in. Any kind of metrics you can share on how much you're willing to extend it? Is it at somewhat of a maximum point, as we exit the year, or is that something that could maybe flex up higher based on market conditions?

是的，謝謝讓我提問。我想問一下 TROI 架構。我相信過去歷史上它少於 12 個月，而面對你們看到的機會，你們似乎願意更積極投入。你們能分享一些指標，說明你們願意把它延長到什麼程度嗎？當我們走出今年時，它是否已經接近某個上限，還是它可能會依市場狀況而再往上彈性調整？

Hey, Josh, it's Nir. So, first of all, this is a framework that allows us to run investment into marketing what we deem is the right cadence and the right risk profile, so to speak on the investment with high discipline. Yes, with very high discipline. Now, to your question, is there a ceiling or a max?

嗨 Josh，我是 Nir。首先，這是一個框架，讓我們能以我們認為合適的節奏與風險輪廓來進行行銷投資，並且在投資上保持高度紀律。是的，非常高度的紀律。那回到你的問題：是否有天花板或上限？

The answer is, right now we think we deem this is to be in a very good framework and we don't expect to increase it necessarily anytime soon. Obviously, many things can change over time that can benefit it, either whether it's going to be more improvement on the gross margin that can be something that's interesting for us.

答案是：目前我們認為這是一個非常好的框架，我們不預期在短期內一定要把它提高。當然，隨著時間推移，很多事情可能會改變並對它有利；例如毛利率進一步改善，這可能會是我們感興趣的因素。

Adaptation of more into annual subscriptions on the Base44 subscription, which will just drive more of the TROI forward, meaning that we collect faster. So there are so many different parts and moving parts throughout this how we calculate this. Then we definitely can make adaptations through it, but it's always through a very clear formula. So it's about us understanding how quickly can we get the investment back.

Base44 訂閱更多採用年繳訂閱，也會把 TROI 往前推進，意思是我們收款更快。所以在我們計算這個指標時，有很多不同的部分與變動因素。那我們確實可以在其中做調整，但永遠是透過一個非常清楚的公式。所以重點在於我們理解：我們能多快把投資收回來。

Okay, and then maybe a follow-up for Lior just on how to think about bookings. If you look to last year, it kind of been within a point of revenue growth and obviously just about 3 points below. Is that a good baseline to kind of use as we model going forward until we maybe start to lap some of these changes? Any pointers on bookings? Or kind of any color on how we should think about modeling partners as well. Thank you.

好的。那再請 Lior 追問一下，關於如何看待 bookings（預訂/簽約額）。如果看去年，它大致在營收成長的 1 個百分點以內，顯然也大約低 3 個百分點左右。在我們建模未來、直到我們開始與這些變化做比較基期（lap）之前，這是否是一個可以採用的良好基準？對 bookings 有什麼建議嗎？或者對於我們應該如何建模合作夥伴端，有沒有一些補充說明？謝謝。

Well, at this point of time, definitely you see in our 2026 situation where you see that revenue growth is have been bookings from obvious reasons, but it doesn't mean that it will continue into 2027. It really depends on how fast we are going to generate more growth, for example, from Base or from other new product that we are going to launch, meaning that if we see that we can get a new product like a new funnel, more customers, generate more growth, Base44 continue under the same rate and even accelerate, for example. So it might be a situation where we see acceleration in bookings. So it's really hard for me to tell you that the same cadence actually will continue into 2027. It might change. I actually hope that it will change.

嗯，在目前這個時間點，你確實會在我們 2026 的情況中看到：由於一些明顯原因，營收成長落後於 bookings，但這不代表它會延續到 2027。這真的取決於我們能多快創造更多成長，例如來自 Base 或我們將推出的其他新產品；也就是說，如果我們看到能用新產品帶來像新的漏斗、更多客戶、帶動更多成長，而 Base44 以相同速度持續甚至加速，例如這樣。那就可能出現我們看到 bookings 加速的情況。所以我很難告訴你同樣的節奏是否會延續到 2027。它可能會改變。我其實希望它會改變。

Okay, very helpful. Thank you.

好的，非常有幫助。謝謝。

Ken Wong, Oppenheimer.

Ken Wong，Oppenheimer。

Fantastic. Thanks for taking my question. I just wanted to kind of dig in on Josh's booking question just now. Any assistance in terms of thinking about that load teams booking trajectory in the second-half? Is it kind of a dip and then reacceleration, fairly consistent both quarters? And then that spread between revenue and bookings of a few points, is that pretty consistent in both 3Q, 4Q?

太好了。謝謝讓我提問。我想再深入追問一下 Josh 剛才關於 bookings 的問題。在下半年，對於 load teams 的 bookings 走勢，你們能否提供一些協助讓我們理解？它會是先下滑再重新加速，還是兩個季度都相對一致？另外，營收與 bookings 之間相差幾個百分點，這在 3Q、4Q 兩季都相當一致嗎？

Hey, Ken, this is Lior. So let's try to understand what is the reason first and then I can answer you. The reason is we spoke about it before about parcels, and we do see a situation where when we are going to deliver more, kind of AI tools to our partners to try obviously to accelerate the growth of partners. We are going to see more partners coming to base.

嗨 Ken，我是 Lior。我們先試著理解原因是什麼，然後我再回答你。原因是我們之前談過的 parcels（聽不清/原文如此），而我們確實看到一種情況：當我們要交付更多、某種程度上是 AI 工具給合作夥伴，以便顯然加速合作夥伴的成長時，我們會看到更多合作夥伴來到 base。

So it's hard for me to tell you about 2027 if it will continue the same way. As I mentioned before, I hope that that won't be the case. But what's happening in the second-half of the year, it's mostly because of partners and obviously we are dealing with it. Right now, the entire company is concentrated only on one thing to generate more profitable growth and we are doing it through base, but also through our new product at Wix, we spoke about it before, more tools to our partners, definitely with Harmony, so I certainly hope that it will change in 2027, but it's too early for me to say.

所以我很難告訴你 2027 是否會以同樣方式延續。如我先前提到的，我希望不會是那樣。但下半年發生的情況，主要是因為合作夥伴，而我們顯然正在處理這件事。目前整家公司只專注在一件事：產生更多具獲利性的成長；我們透過 base 來做，也透過我們在 Wix 的新產品來做——我們之前談過——提供更多工具給合作夥伴，當然也包括 Harmony；所以我確實希望 2027 會改變，但現在說還太早。

Okay, understood. And then just on the revenue guide, the implied Q4 number does seem to suggest a bit of acceleration there and the comp is still pretty tough and the bookings have been a little soft in the first half. So just the confidence in that 4Q revenue number, what's underpinning that, Lior?

好的，了解。另外就營收指引來看，隱含的第四季數字看起來確實暗示有些加速，但比較基期仍然相當艱難，而且上半年預訂（bookings）有點偏軟。所以對第四季營收數字的信心來源是什麼？Lior，是哪些因素在支撐？

So I think that the Q4 revenue, first of all, is I don't see it much different than the Q3, meaning that I don't see acceleration in Q4 compared to Q3, but definitely we see a lot of benefit coming mostly from Base44 into the numbers. So Q3 and Q4 revenue growth on a year-over-year basis more or less the same, but I don't see acceleration in Q4 compared to Q3.

我認為第四季營收，首先，我不覺得它和第三季有太大不同，也就是說，我不認為第四季相較第三季會加速，但我們確實看到主要來自 Base44 的貢獻會在數字上帶來很多好處。因此，以年對年來看，第三季與第四季的營收成長率大致相同，但我不認為第四季會比第三季加速。

Okay, perfect. Thank you.

好的，完美。謝謝。

Robert Coolbrith, Evercore.

Robert Coolbrith，Evercore。

Hi, two questions please. As partners use Base44 in place of Wix Studio, just wondering if you could talk about attach rate of Wix Business Solutions on a headless basis. Maybe also just some early learnings from some of the headless initiatives you had with the LLM partners as well.

嗨，兩個問題，謝謝。當合作夥伴用 Base44 取代 Wix Studio 時，想請你談談在 headless 架構下 Wix Business Solutions 的附加率（attach rate）。另外也想聽聽你們與 LLM 合作夥伴推動的一些 headless 計畫的早期學習與觀察。

And then just wanted to double click on the GM profile in H2. I think you said one to two points of benefit, H2 over H1 from Base44 or the Base 1 model specifically. But then you also talked about some AI savings in the rest of the business. So just wondering if you were to put that all together, maybe you could give us a consolidated view of the GM improvement in H2. Thank you.

另外我想再深入追問一下下半年（H2）的毛利率（GM）輪廓。我記得你說下半年相較上半年（H1），Base44 或更具體說 Base 1 模型會帶來 1 到 2 個點的利多。但你也提到在其他業務還有一些 AI 節省。所以想請問如果把這些都合在一起，你能否給我們一個下半年毛利率改善的整體（consolidated）視角？謝謝。

Yeah. I will start with the gross margin and about the two points that you just mentioned. And then Avishai will continue with the first one on (inaudible).

好的。我先從毛利率以及你剛提到的那 2 個點開始說。然後 Avishai 會接著回答第一個問題，關於(聽不清楚)。

So with regard to the gross margin improvement in the second-half, I said that it's going to improve by 2 points and this is mostly coming from the savings that we see compared to the guidance we provided previously about Base44, meaning that the improvement in gross margin and the usage of the model of Base 1.

關於下半年毛利率的改善，我說會提升 2 個點，這主要來自於我們相較先前對 Base44 所提供指引所看到的節省，也就是毛利率的改善，以及 Base 1 模型的使用。

And the fact that it's, as Avishai mentioned, that it even performed better, so we see that the usage of Base 1 is even much higher and better than what we expected and faster. Therefore, we updated the gross margin as a result of that, meaning that we see a dramatic change in gross margin of Base44 and it's translated to a 2-point improvement in the consolidated revenue, in the consolidated gross margin.

而且如 Avishai 提到的，它的表現甚至更好，所以我們看到 Base 1 的使用率比預期更高、更好，而且導入更快。因此，我們據此更新了毛利率；也就是說，我們看到 Base44 的毛利率出現顯著變化，並轉化為合併毛利率提升 2 個點（在合併層面）。

Hey, Robert, for the first part of your question, so as Avishai mentioned before, partners, you asked about partners using Base44 instead of studio, the dynamics here, as Avishai said about partners and Base44 are very early.

嗨，Robert，關於你問題的第一部分，如 Avishai 先前提到的，你問到合作夥伴用 Base44 取代 Studio 的情況；這裡的動態，如 Avishai 所說，合作夥伴與 Base44 目前都還非常早期。

And it's very -- I think it's too early to comment about whether it's supplementing or it's actually additional. By the way, we are seeing some that are using both for different use cases. And I think that as time progresses, we'll probably have segmentation more and more of some partners and agencies that are sticking to one product, other that are using different products for different kind of projects that they're doing.

而且這非常——我認為現在評論它究竟是在補充（supplementing）還是其實是額外增量（additional）都還太早。順帶一提，我們確實看到有些人會針對不同使用情境同時使用兩者。我想隨著時間推進，我們可能會看到更明確的分眾：有些合作夥伴與代理商會固定使用某一個產品，另一些則會針對不同類型的專案使用不同產品。

So I think it's early in this cycle, it's definitely an evolution of this segment and it should be very interesting. Our goal is to try to be there and deliver the right value on whether they're trying to go one way or the other. By the way, to some extent, and the second part of your question was about headless, it also applies to headless.

所以我認為在這個週期的早期階段，這個族群肯定正在演進，會非常有意思。我們的目標是要在那裡，無論他們想走哪一條路，都能提供正確的價值。另外，你問題的第二部分是 headless；某種程度上，這也同樣適用於 headless。

Obviously, in a world where there are more and more agentic solutions out there, we want people to be able to benefit from the value of the business stack that's on Wix in a very easy manner, and this is what is headless is all about. It's something we're starting to see pick up, but again, it's an early cycle. Obviously, it's much, it is aimed toward people who are more professional in their youth and what they're trying to build.

顯然，在一個越來越多代理式（agentic）解決方案出現的世界裡，我們希望大家能以非常容易的方式，從 Wix 上的商業技術堆疊（business stack）中獲得價值，而這正是 headless 的核心。我們開始看到它有起色，但同樣地，仍處於早期週期。很明顯，它更是面向更專業的使用者，以及他們想要打造的東西。

And we're seeing traffic coming to this from various places over the Internet, where the professionals are dealing and working with agentic solutions, it can be things that are coming from a topic, OpenAI, our own stack. And our goal is to definitely be there and add that to the portfolio of solutions that we offer at the professional clouds. I do think that in the coming few quarters, it's going to become more and more clear exactly what is the preferred path for each and every one of them.

我們也看到來自網路各處的流量導入到這裡，因為專業人士正在接觸並使用代理式解決方案；來源可能是某個主題、OpenAI、或我們自己的技術堆疊。我們的目標是一定要在那裡，並把它加入我們在專業雲端（professional clouds）所提供的解決方案組合中。我確實認為在接下來幾個季度，對於每一類使用者而言，哪一條路徑是偏好的，會變得越來越清楚。

Great. Thank you.

很好。謝謝。

Naveed Khan, B. Riley Securities.

Naveed Khan，B. Riley Securities。

Thank you very much. I understand you guys are not updating the ARR for Base44, but just curious if the trajectory is similar to what we saw between your last few updates. And are you still taking share from the level in the US? Are you still ahead there because I think that's something you pointed to before. So that's my first question.

非常感謝。我了解你們不會更新 Base44 的 ARR，但我想知道其走勢是否與你們前幾次更新時看到的情況相近。另外，你們在美國是否仍在從 Lovable 那邊取得市占？你們在那邊是否仍然領先？因為我記得你們之前有提到這點。這是我的第一個問題。

And then, in terms of just say Base44 mix of monthlies versus annuals, can you just maybe talk about how that mix looks like? It seems like you're getting more annuals versus monthlies, but any color that would be helpful. Thank you.

第二個問題是，就 Base44 的月繳（monthly）與年繳（annual）組合來看，你能否談談目前大概的結構？看起來你們年繳比月繳更多，但如果能提供一些補充說明會很有幫助。謝謝。

Hi, Naveed. So for the ARR, as we said before, this is not a KPI we intend to report. It's trending in similar manners as before. In terms of what you asked about the mix of monthly and annual, then, yes, I would say there is more adoption of annual, but we are not in a place where we want to break it out to the exact numbers right now.

嗨，Naveed。關於 ARR，如我們之前所說，這不是我們打算揭露的 KPI。它的趨勢與之前相似。至於你問到月繳與年繳的組合，是的，我會說年繳的採用更多，但我們目前不打算把精確數字拆分揭露。

For the Lovable market share in the US, you want to comment on that, Adishai?

至於在美國的 Lovable 市占，你想評論一下嗎，Adishai？

Yes, I think that it's important to say we are not taking shares from Lovable, we are inventing. A new market together with Lovable. And we are competing on the percentage of that going forward. But I think it's mostly us supporting each other more than competing with each other by educating the market and educating people that the fact that they can now build applications themselves and invent products and actually make that manifest that in reality, which was shocking for most people. So we, I would say we more support each other than competing with each other.

是的，我認為重要的是要說，我們不是在從 Lovable 那裡搶市占，我們是在「發明」——與 Lovable 一起創造一個新市場。而我們是在競爭未來這個市場中各自所占的比例。但我認為更多是彼此支持，而不是彼此競爭：透過教育市場、教育大眾，讓大家知道他們現在可以自己打造應用程式、發明產品，並且把它真正落地到現實中——這對大多數人來說是很震撼的。所以我會說我們彼此支持多於彼此競爭。

Yeah, thank you guys. Yeah, I misspoke. I think you have said that you were ahead of Lovable in terms of taking the share, but not necessarily.

好的，謝謝你們。對，我剛剛說錯了。我想你們之前是說你們在取得市占方面領先 Lovable，但不一定是那個意思。

But what I think is very hard if you're trying to find the precise market split between us to Lovable, it's extremely hard. They are a private company. They report numbers based on metrics that are not clear. Okay, so it's very hard for us to distinguish marketing on the numbers from real numbers.

但我認為，如果你想要找出我們與 Lovable 之間精確的市場分配比例，這非常困難。他們是一家私人公司。他們用一些不清楚的指標來公布數字。所以對我們來說，很難分辨行銷上的數字與真實數字之間的差異。

And I can say that our estimation is that in the more important market, the United States, we are ahead of them. But to say by how much is again tricky. We can only look at secondary metrics that we can measure on the Internet. I would say we're in a very good place.

我可以說，我們的估計是，在更重要的市場——美國——我們領先他們。但要說領先多少又很棘手。我們只能看一些我們能在網路上衡量的次級指標。我會說我們處於非常好的位置。

Great. Thank you, guys.

很好。謝謝各位。

Andrew Boone, Citizens.

Andrew Boone，Citizens。

Thanks so much for taking the questions. I wanted to go back to Base 1. Can you just talk about the improvements of the model? If we think about the trajectory of Base 1, should we really think about a parallel path with open source more broadly? So as we think about Kimi K3, if we think about other open source models that have come to market, is that the right trajectory of what we see for improvements for Base 1 or how do we think about that? And then just connect that more broadly to what we're seeing in terms of conversion or retention, how do we think about the improvements of models with Base44 improving the overall business?

非常感謝接受提問。我想回到 Base 1。你能談談模型的改進嗎？如果我們思考 Base 1 的發展軌跡，我們是否應該把它視為與更廣泛的開源路徑平行前進？所以當我們看 Kimi K3、以及其他已經上市的開源模型時，這是否就是我們對 Base 1 改進所看到的正確軌跡？或者我們應該如何看待？另外，把這點更廣泛地連結到我們在轉化或留存方面看到的情況：當 Base44 改善模型時，我們如何看待 Base 1 的改進對整體業務的提升？

And then secondly, just more of a bigger picture question. A lot of people think about frontier models and their progress and especially just the competitive threat from Claude Code or other frontier type models. How do we think about just the competitive moat, broadly speaking, for Base 1 and open source more broadly versus what is the frontier as we think specifically about cutting assistance. Thanks so much.

第二個是更宏觀的問題。很多人會思考前沿模型及其進展，尤其是來自 Claude Code 或其他前沿類模型的競爭威脅。就廣義而言，當我們特別聚焦在程式碼輔助時，我們如何看待 Base 1 與更廣泛的開源相對於前沿模型的競爭護城河？非常感謝。

All right, so those are three different questions. If what one is Base 1, would it improve? In parallel with open source model. So I want to be saying, first of all, the answer is yes, of course. However, I don't think that is the most significant part of the improvement.

好，這是三個不同的問題。第一個是 Base 1 會如何改進？是否會與開源模型並行改進。我想先說，答案當然是會。不過，我不認為那是改進中最重要的部分。

For most of the applications that you want to build today, Kimi 3, GLM 5.2, 5.4, I think, and Claude [or 5], or the latest codecs, they are all good enough for most applications. Because most of us don't do those complicated things that require that extra few percent in model intelligence. So for most of the applications, the improvement that we need to make on Base 1 are to build those applications better. And it's a different thing, it's not about how we make it more intelligent in Hungarian poetry, right, which is one of the things that, or on any other of those things, molecular biology, right, all those things that those models are trying to push the envelope.

對於你今天想要打造的大多數應用來說，Kimi 3、GLM 5.2、5.4（我想是）、以及 Claude [或 5]，或最新的 codecs，它們對大多數應用都已經夠好了。因為我們大多數人不會做那些需要模型智慧再多那幾個百分點的複雜事情。所以對大多數應用而言，我們需要在 Base 1 上做的改進，是把那些應用做得更好。這是不同的事情；不是在於如何讓它在匈牙利詩歌之類的事情上更聰明，對吧——那是其中一種，或任何其他事情，例如分子生物學，對吧——那些模型正在嘗試突破極限的領域。

A lot of what we need to do is how we make applications that people prompt, and a lot of times those people are not professional developers, they don't prompt them in a professional way, right, they're lacking a lot of the definitions, they don't put all the information that you need, how do you make base.

我們需要做的很多事情，是如何把人們用提示（prompt）來操作的應用做得更好；而很多時候，那些人並不是專業開發者，他們不會用專業的方式下提示，對吧，他們缺少很多定義，也沒有放入你所需要的全部資訊；我們要如何讓 base。

I'll make Base 1 better than solving those specific problems. And so, yes, we will benefit from continuation of development in the open source environment, absolutely. But the vast majority of the value will come from what we do and not from implementing the open source models.

我會讓 Base 1 透過解決那些特定問題而變得更好。所以，是的，我們絕對會受益於開源環境中持續的發展。但絕大多數的價值會來自我們自己做的事情，而不是來自於導入開源模型。

How does Base 1 improve impact conversion and retention? I believe that was your second question. Well, obviously, if you come to Base44 and you prompt something and you get a bad application, you're going to convert a lot less than if you went to Base44 type the same prompt and get a better application.

Base 1 的改進如何影響轉化與留存？我相信那是你的第二個問題。很顯然，如果你來到 Base44，輸入一個提示，結果得到一個很差的應用，那你的轉化率會遠低於你在 Base44 輸入同樣提示卻得到更好應用的情況。

Even if you don't get a pivot application in a place that you can now feel confident that if you start working and continuously to prompt, you'll get to what you want, that's already a dramatic improvement in your chances to convert. And we see that, we measure that, we have now a year of consistent measurements on that.

即使你沒有立刻得到一個「完美」的應用，但只要到了一個你能有信心的程度——也就是你開始動手、並持續透過提示迭代，就能走到你想要的結果——那就已經會大幅提升你轉化的機率。我們看得到這點，我們也有量測；我們現在已經有一年一致性的量測數據。

We know that that's the case. So by improving Base 1, I'm confident we will continuously improve Base44 conversion, which is why I'm so excited about it. The other side of it, of course, is exactly the same for retention. What is the competitive moat for Base 1 and open-source broadly based on [Tmalls]?

我們知道確實如此。因此，透過改進 Base 1，我有信心我們會持續提升 Base44 的轉化，這也是我對此如此興奮的原因。當然，另一面對留存也是完全一樣。那麼，Base 1 與更廣泛的開源，其競爭護城河是什麼，基於[Tmalls]？

That's a very long question. I think that it's beyond the scope of this conversation. I'll just say my $0.02. Currently, everybody is using exactly the same algorithm. Okay, yes, we do some engineering modification on top of it, and but that's pretty much it, so the mod is not huge, okay, for frontier models.

這是一個很長的問題。我認為這超出了這次對話的範圍。我只說說我的一點看法。目前，大家用的基本上是同一套演算法。好，沒錯，我們會在上面做一些工程上的修改，但大致也就如此，所以對前沿模型而言，護城河並不大。

However, they've proven in the past that they have the ability to innovate even within the same algorithm, which is attention-based transformers, right? That's the algorithm, and create new things. So if they continue to do that, they'll have some kind of remote. There is another side to that is that there is a level where it doesn't matter or legal action or government action will start preventing the evolution and release of new models. We see that we meet this right on security, cyber security issues.

不過，他們過去已經證明，即使在同一套演算法之內——也就是基於注意力的 Transformer，對吧？那就是演算法——他們仍然有能力創新並做出新東西。所以如果他們持續這麼做，他們就會有某種護城河。另一方面，也存在一個層級：到了那個程度，這就不再重要，或者法律行動、政府行動會開始阻止新模型的演進與發布。我們在安全、資安議題上正好就遇到這種情況。

So I think that between, and the fact that for a lot of the times, the current model are already intelligent enough. So if you're a commercial model, a continual commercial model, but for the task that your customers need, that extra intelligence that you've added are not significant. It's very hard to justify prices.

所以我認為，在這些因素之間，再加上很多時候現有模型其實已經足夠聰明。因此，如果你是商業模型、持續商業化的模型，但對客戶所需的任務而言，你額外增加的那點智慧並不顯著。那就很難為價格找到合理性。

In other words, it's a very interesting conversation, I think beyond the conversation of our discussion, but, I think it's extremely important for the Western world to have really strong frontier models in the Western world. And I really hope that those guys at OpenAI and Entropic and Google will continue to innovate and drive fantastic products.

換句話說，這是一個非常有意思的話題，我想已經超出我們這次討論的範圍；但我認為，西方世界擁有非常強的前沿模型極其重要。我也真心希望 OpenAI、Entropic 和 Google 的那些團隊能持續創新，並推出很棒的產品。

Thank you.

謝謝。

Thank you very much. This concludes our question-and-answer session. Thank you for your participation in today's conference. You may now disconnect.

非常感謝。我們的問答環節到此結束。感謝各位參與今天的會議。您現在可以斷線。