
2022年,电子纸单词卡这一新品类在K12教育硬件市场中悄然诞生,首款产品10个月即卖出80万台。
四年后的今天,电子纸单词卡成为2400亿元教育智能学习硬件市场里跑出来的“小而美”代表,市场规模连年大涨,成为K12学生、考研留学、记忆学习的主流工具。
根据洛图科技(RUNTO)最新发布的2026年Q1数据:
2026年第一季度,电子纸单词卡销量20.6万台,同比再增45.6%,销额6122.7万元,增速80.4%。洛图科技预测,2026年上半年销量将达44万台,增长46%。
在过去的2025年全年,电子纸单词卡的销量与销额均达新高。线上销量达81万台,同比增长66.6%;销额2.2亿元,同比增长114.2%。
从2022年的约20万台,到2024年的近50万台,再到2025年的81万台——3年翻了4倍。
但比这些数字更耐人寻味的,是市场主角的变化。
当年仅用常规项目20%资源孵化喵喵机单词卡的产品开创者作业帮,正在悄悄转移战场。2026年Q1,其旗下喵喵机的市占率已不足2%。同一份报告里,另一家公司的名字却格外扎眼——新东方,稳居前三,快速增长。
而统治这个市场的,是另一个靠10年内容积累筑起护城河的玩家——百词斩。2026年Q1,它只靠一款机型就拿走了市场65.5%的份额。

缝隙里的生意:
销售81万台,只是单词卡的起点
单词卡正在悄悄“蚕食”词典笔的市场。
虽然单词卡81万台的年销量,远低于词典笔的530万台,但其增速和潜力恰恰相反。
大量词典笔的购买用户只会高频使用其中10%的功能,即“查词和背词”,而这两个功能,一台300元的单词卡就能覆盖。
用户的实际决策路径正在从一步到位买词典笔变成先买台单词卡试试。这个"降级替代"的逻辑一旦成立,单词卡的上限就不止81万台了。
单词卡的竞争力,源于它解决了一个真问题:心无旁骛地随时随地背单词,它找到了一个手机做不到、词典笔太贵的精准切口。
3亿人学英语,约5000万人有背单词和购买工具的需求。但手机App诱惑太多,打开下一秒可能就在刷抖音;纸质书不方便随身携带;词典笔上千元,核心是"查"而非"背"。
单词卡专注于做减法,卡在让用户一心一意地背单词的缝隙里。
市场也在给出积极信号,从洛图科技给出的单词卡全年数据来看,单词卡均价正在往上走。300-499元价格段的产品跃升为销售主力,销售份额占比过半。
该价格段2024年销量仅为5.7万台,2025年增至40.6万台,同比增幅高达606.2%,市场占比为50.1%。
与之形成鲜明对比的是,曾经的第一价格段100-199元,在2025年的销量下滑至16.2万台,同比降幅为19.1%。此外,0-99元的超低价段的销量亦同比大幅下滑35.7%,降至2.3万台。

价格上涨,用户仍愿意买单的原因,是单词卡的产品价值有了实质性的提升。
单词卡早已不是那个“只能背英语单词”的黑白屏小玩意。产品内容大幅扩容,不止英语,新增日语、韩语、多语种;不止语言,加入语文古诗词、数学公式定理、法律条令;
同时新增中高考真题收录、高频词汇标识等硬核功能,还通过AI测评生成单词水平表,适配对应的学习方案。
用户群体从K12扩展到大学生、考研党、职场考证人群;功能从单一记忆到多功能测试、AI辅助记忆等,产品的天花板被明显抬高。

195个品牌抢市场
前5家拿走约98%份额
需要看清的是,这并非一个百花齐放的赛道,至少目前看来这是一场赢家通吃的残酷牌局。
2022年上半年,这个赛道只有38个品牌。
作业帮的喵喵机第一个吃螃蟹后,网易有道、外研通、名校堂、高途、途蛙、新东方、联想、科大讯飞——教培的、硬件的、互联网的,全都来了。
到2025年底,线上渠道在售的电子纸单词卡品牌,增加到了195个。
195个品牌抢一个81万台的市场,平均每个品牌年销4150台,月均不到350台。这是一个看起来"谁都能做、谁都不赚钱"的赛道。
洛图科技2026年Q1的数据,揭开了一个真相:百词斩、新东方、科学罐头、科大讯飞、喵喵机前5家拿走了约98%的销量。剩下的190家,分享不到2%。

这张牌桌上,三个名字勾勒出了品类的命运曲线。
先看在单词卡市场中一骑绝尘、一家独大的百词斩。2026Q1数据显示,在前5家拿走近98%的销量,而百词斩仅靠MAX一款机型就拿走了65.5%的份额。
百词斩做语言学习App超过10年,积累了完整的词库生态、记忆算法和数千万用户的学习数据。当别人还在采购公版词书时,百词斩已经把自家App上的词库、算法、用户画像完整迁移到了硬件上。
但真正让它在年轻用户中“封神”的,不是算法,而是对Z世代审美和语境的精准拿捏。
2025年底,百词斩推出与Chiikawa(吉伊卡哇)的联名单词机,三款礼盒上线即售罄。Chiikawa作为当下Z世代备受追捧的IP之一,联名款自带社交话题属性——学生买的不只是一台单词机,更是一个“我在潮流中”的身份标签。
与此同时,百词斩单词机MAX还植入了“宠物养成”游戏机制。用户完成每日词汇任务,就能解锁宠物角色的成长地图。这种“养成系”设计把枯燥的背单词变成了“养电子宠物”,显著提升了学习粘性和用户粘性。
其二,作为电子纸单词卡,曾经的开创者作业帮,如今已经成为“边缘人”,份额从第一掉到了第五。
2022年,喵喵机单词卡市占率51.4%——每卖出两台,就有一台是喵喵机。四年后,其市占率已不足2%。不是被打败,是主动退场。
当公司战略重心转向学习机、学习桌等更高客单价的硬件时,单词卡这个“轻项目”被边缘化。与此同时,单词卡的竞争逻辑从护眼、续航转向内容生态和AI算法,作业帮的“题库基因”已不再适配。
其三,新东方通过教研资产复用和产品线稳步迭代,增速跑赢大盘。
与作业帮的“撤退”形成对比,新东方瞄准了两件事:
一是教研资产直接注入硬件。新东方的词库不是采购的,而是30年教学积累的“教研词库”——每一个单词的释义、例句、记忆方法,背后都是新东方老师反复打磨过的。
二是多产品线并行,价格卡位灵活。单词通专注英语背词,背词宝扩展到九大学科;纪念版99.9元做引流,Pro版冲击中高端。结果是新东方在单词卡市场份额占比进入前三。
有些品牌选择了一条截然不同的路,不跟巨头正面竞争,而是找到自己的生存空间。
科学罐头走了一条“向下兼容”的路线,当所有人都在抢中学生和大学生时,它用"自然拼读+实体键盘+AI游戏化"打低龄英语启蒙市场。产品叫"自然拼读口袋单词机"。
不是传统的"背单词"概念,而是"学发音+拼写+游戏PK",切的是幼小衔接的刚需,仅用一个寒假就变成了低龄爆款。
科大讯飞的差异化,写在了自己的DNA里:AI语音。当其他单词卡还在显示单词和释义时,讯飞的产品已经做到了智能口语测评、AI发音纠正。
这需要的是语音识别和自然语言处理的底层技术积累。讯飞不拼词库数量和联名IP,它在争夺的是那些"不仅要记住单词,还要能说出来"的进阶用户。

大模型时代
单词卡急需一个“新故事”
值得注意的是,2023-2025年,中国单词卡相关技术专利申请量年均增长超120%,集中在自适应学习算法、语音交互与评测、脑认知科学应用等方面。
AI大模型使得设备能够实现更精准的艾宾浩斯记忆曲线规划、上下文语义化例句生成以及动态难度调整。
技术狂奔的同时,一个根本问题浮出水面:当每家品牌都能做出“聪明的、个性化的”的单词卡,用户为什么选你?
从全球教育AI的发展趋势来看,真正的分水岭正在形成。以Quizlet为代表的老牌巨头通过AI诊断、AI自动规划与AI自适应测试,模拟真实考试场景等功能,月活用户突破2055万。
而以多邻国为代表的产品,则将游戏化机制与GPT-4深度结合,Q4月均营收推高至2亿美元,证明了“有趣+智能”才是真正的黏性引擎。
上海外国语大学教授梅德明对此有一个形象的比喻:“单词就像是一片片树叶,树叶要在树上才有生命力。如果只学词汇,就像树叶会被风吹走。”
他在接受采访时强调,脱离语言环境的孤立记忆不是有效的语言学习。词汇学习的旨归,是在真实语境中运用词汇表达思想、解决问题。
这个判断,恰好点中了单词卡品类的命门。
过去的单词卡解决的是记住的问题。但当AI让个性化记单词变得人人可及,下一个战场必然是如何用起来。
目前,海外市场已出现探索性案例。AI语言学习平台Speak通过AI Tutor构建真实对话场景,帮助用户完成语言输入输出一体化。
丹麦初创公司Alice.Tech聚焦应试词汇学习,用AI将课程转化为个性化词汇卡片与练习模块进行记忆,2025年完成480万美元种子轮融资。
这意味着单词卡的竞争正在从记忆工具升级为更深层的语言运用能力的较量。
回到国内市场,百词斩的答案一直是内容生态,新东方的答案是教研体系,科学罐头的答案是游戏化启蒙——但这些都不是AI技术本身给的。
当技术变成通用能力,单词卡的下一场战争,不在技术里,而在谁能帮用户把单词用起来的命题里。
作者:玉龙
头图来源于网络

In 2022, the e-paper word card, a new product category, quietly emerged in the K12 educational hardware market, with the first product selling 800,000 units in just 10 months.
Four years later, the e-paper word card has become a representative "small yet beautiful" success story in the $24 billion intelligent learning hardware market, with market size growing significantly year after year, making it the mainstream tool for K12 students, graduate exam candidates, and memory learners.
According to the latest data released by RUNTO Technology for 2026 Q1:
In the first quarter of 2026, e-paper word card sales reached 206,000 units, a 45.6% year-over-year increase, with a sales value of 61.23 million RMB, up 80.4%. RUNTO predicts that sales in the first half of 2026 will reach 440,000 units, representing a 46% growth.
For the full year of 2025, both the sales volume and value of e-paper word cards reached new highs. Online sales hit 810,000 units, up 66.6% year-over-year; sales value was 220 million RMB, up 114.2%.
From approximately 200,000 units in 2022, to nearly 500,000 in 2024, and then to 810,000 in 2025 — the market has quadrupled in three years.
However, more intriguing than these numbers are the shifts in the market's main players.
The product creator, Zuoyebang, which incubated the Miaomiao Word Card with only 20% of its regular project resources, is quietly shifting its focus. By 2026 Q1, its Miaomiao series held a market share of less than 2%. In the same report, another name stands out — New Oriental, firmly ranking in the top three with rapid growth.
But the true ruler of this market is a player that built its moat over 10 years of content accumulation — Baicizhan. In 2026 Q1, it captured 65.5% of the market with just one model.

A Niche Business:
810,000 Units Sold is Just the Beginning for Word Cards
Word cards are quietly "eroding" the market for electronic dictionaries.
Although the annual sales of word cards (810,000) are far lower than those of electronic dictionaries (5.3 million), their growth rate and potential are inversely proportional.
The majority of electronic dictionary buyers only heavily use 10% of the features — namely, "word lookup and memorization" — which a 300 RMB word card can fully cover.
The actual decision-making path for users is shifting from purchasing an all-in-one electronic dictionary to first trying a word card. Once this "downgrade replacement" logic takes hold, the ceiling for word cards is not limited to 810,000 units.
The competitiveness of word cards stems from their ability to solve a real problem: allowing for uninterrupted, anytime-anywhere word memorization. They have found a precise niche that smartphones cannot fill and electronic dictionaries are too expensive for.
With 300 million people learning English, approximately 50 million have the need to memorize words and purchase tools. However, smartphone apps are full of distractions (like TikTok), physical books are inconvenient to carry, and electronic dictionaries are primarily for "lookup," not "memorization," costing over 1,000 RMB.
Word cards excel by subtracting distractions, fitting perfectly into the niche of enabling users to focus solely on memorization.
The market is responding positively. According to RUNTO's full-year data, the average price of word cards is rising. Products in the 300–499 RMB price range have become the sales leaders, accounting for over half of the market share.
In 2024, sales in this price segment were only 57,000 units; in 2025, they skyrocketed to 406,000 units, a 606.2% year-over-year increase, capturing 50.1% of the market.
Conversely, the once-dominant 100–199 RMB segment saw sales drop to 162,000 units in 2025, a 19.1% year-over-year decline. The ultra-low-price 0–99 RMB segment also fell significantly by 35.7% to 23,000 units.

The reason users are willing to pay higher prices is that the value of word card products has substantially increased.
Word cards are no longer just "black-and-white screens for memorizing English." Their content has expanded significantly, now including Japanese, Korean, and multiple languages, as well as subjects like Chinese poetry, mathematical formulas, and legal provisions.
They have also added high-value features such as recording actual exam questions, identifying high-frequency vocabulary, and using AI to generate personalized learning plans based on user levels.
The user base has expanded from K12 students to university students, graduate candidates, and professionals, while functionality has evolved from simple memory storage to complex testing and AI-assisted learning, significantly raising the product's ceiling.

195 Brands Contest the Market
The Top 5 Capture ~98% of the Share
It is crucial to understand that this is not a market of diverse players; rather, it appears to be a winner-takes-all game.
In the first half of 2022, only 38 brands operated in this space.
After Zuoyebang's Miaomiao pioneered the category, giants like NetEase Youdao, Waiyan Tong, Mingxiao Tang, Gaotu, Tufrog, New Oriental, Lenovo, and iFLYTEK all rushed in — encompassing both education and hardware sectors.
By the end of 2025, the number of electronic word card brands selling online had grown to 195.
With 195 brands dividing an 810,000-unit market, the average annual sales per brand is only 4,150 units, or less than 350 units per month. This suggests a market where "anyone can enter, but no one makes money."
RUNTO's 2026 Q1 data reveals the truth: the top 5 brands (Baicizhan, New Oriental, Scientific Can, iFLYTEK, and Miaomiao) account for approximately 98% of sales. The remaining 190 brands share less than 2% of the market.

On this playing field, three names define the trajectory of the industry.
First, consider Baicizhan, the undisputed leader in the word card market. According to 2026 Q1 data, though the top 5 capture nearly 98% of sales, Baicizhan alone claims 65.5% with just a single model (MAX).
Baicizhan has been developing language learning apps for over 10 years, accumulating a complete ecosystem of vocabulary libraries, memory algorithms, and user behavior data. While others were purchasing generic word lists, Baicizhan had already migrated its entire app-based curriculum to hardware.
But what truly sets it apart from younger users is not the algorithm, but the precise understanding of Gen Z aesthetics and culture.
At the end of 2025, Baicizhan launched a limited-edition word card collaboration with Chiikawa, a beloved IP among Gen Z. The three gift sets sold out immediately. This partnership leverages the cultural currency of Chiikawa, making the purchase not just a learning tool, but a statement of trendiness and identity.
Additionally, the Baicizhan MAX card incorporates "pet breeding" game mechanics, where users unlock character growth maps by completing daily vocabulary tasks. This "gamified" approach transforms boring memorization into fun, significantly improving user stickiness.
Second, is Zuoyebang, the pioneer of the word card category, which has now become a "fringe player," dropping from first place to fifth in market share.
In 2022, Miaomiao Word Cards held a 51.4% market share — every other card purchased was likely a Miaomiao. Four years later, its share is below 2%. This is not a story of defeat, but of strategic withdrawal.
As Zuoyebang shifted its corporate focus to higher-margin products like learning machines and desks, the "light" word card project was neglected. Furthermore, the nature of competition shifted from battery life and display to content ecology and AI, areas where Zuoyebang's problem-book focus was insufficient.
Third, is New Oriental, which has successfully grown its share by leveraging its rich educational assets and diversifying its product line.
Unlike Zuoyebang, New Oriental focused on two main areas:
First, directly injecting its educational assets into the hardware. New Oriental's vocabulary database is not third-party; it is a "curriculum-based dictionary" refined over 30 years of teaching, with every definition and example sentence scrutinized by experienced educators.
Second, maintaining a flexible product line. "Word Master" caters to English memorization, while "Memory Card" covers nine additional subjects. They also offer a 99.9 RMB commemorative edition for entry-level users and a Pro version for the high-end market. This strategy has propelled New Oriental into the top 3.
Some brands have chosen a different path, avoiding direct competition with giants by finding their own niche.
For example, Scientific Can targeted the "downstream" market of young learners. While everyone else focused on middle and high school students, they created the "Phonics Pocket Word Card" centered around phonics, a physical keyboard, and AI games. This product addressed the specific needs of early learners (kindergarten to grade 2) and became a winter break hit.
iFLYTEK, meanwhile, leverages its unique DNA in AI and speech recognition. Unlike standard cards that simply display words, iFLYTEK offers intelligent speech evaluation and AI pronunciation correction. This capability transforms the device into a speech trainer rather than just a memory aid, appealing to users who want to speak as well as read.

The Age of Large Language Models:

It is worth noting that between 2023 and 2025, patent applications related to word cards grew by over 120% annually, focusing on adaptive learning algorithms, voice interaction/evaluation, and brain science applications.
Large AI models (LLMs) have enabled devices to perform more precise spaced repetition (based on the Ebbinghaus curve), generate context-rich example sentences, and dynamically adjust difficulty levels.
However, as technology races forward, a fundamental question emerges: if every brand can create a "smart, personalized" word card, what sets one product apart?
Looking at global education trends, the true differentiator is not just memorization, but usage. Platforms like Duolingo ($200M monthly revenue) and Quizlet (20.55M monthly active users) have proven that integrating AI with practical application (games, quizzes, or conversation) leads to higher engagement.
Professor Mai Deming from SISU aptly states, "Words are like leaves; they only have life on a tree. If you only learn words, they are like leaves that can be blown away by the wind."
The future of word cards lies not in how much vocabulary you can fit onto the flashcard, but how effectively that vocabulary can be used to communicate.
Currently, overseas markets are exploring this with companies like Speak (AI Tutor) and Alice.Tech (exam-focused AI generation).
Back in China, while Baicizhan focuses on ecology and New Oriental on curriculum, none of these are purely technological advantages specific to the hardware. The next phase of competition will require brands to help users transition from passive memorization to active language use.
In his interview, he emphasized that isolated memory without a language environment is not effective language learning. The goal of vocabulary learning is to use words to express ideas and solve problems in real-world contexts.
This insight strikes at the core of the flashcard market.
Previously, flashcards solved the problem of remembering. But now that AI makes personalized memorization accessible to everyone, the next battle will be about how to use this knowledge effectively.
Overseas, experimental models have emerged. The AI language learning platform Speak uses AI tutors to create realistic dialogue scenarios, helping users integrate language input and output.
The Danish startup Alice.Tech focuses on exam-oriented vocabulary learning, transforming courses into personalized flashcards and practice modules, securing $4.8 million in seed funding in 2025.
This indicates that the competition for flashcards is evolving from a memory tool race to a contest over deeper language application abilities.
In the domestic market, Baicc's answer has been content ecosystem, New Oriental's is teaching and research system, and Science Can's is game-based enlightenment—but none of these stem from AI technology itself.
As technology becomes a common capability, the next war for flashcards will not be fought in the realm of technology, but in the mission of helping users apply their words effectively.
作者:玉龙
头图来源于网络