各位法友新年好!
承蒙您一路的護持與支持,CBETA 即將邁向第 28 年電子佛典化的工作。
回顧 2025 年,我們持續以穩健踏實的步伐,推進佛典數位化與長期保存的各項工作,並在既有基礎上,逐步累積可被檢驗、可被使用的具體成果。
這一年,我們一方面持續擴充佛典資料庫內容,深化經文整理、校訂與新式標點等基礎工程;另一方面,也審慎地嘗試將新技術導入實際作業流程,期望在不影響文本品質與學術可靠性的前提下,提升整體處理效率與使用體驗。
在上述工作的基礎上,2025 年 CBETA 亦將相關研究成果帶入國際學術交流場域。同年 10 月,CBETA 籌組「Reimagining Buddhist Scholarship with Artificial Intelligence(人工智慧時代的佛教學術轉型)」的論文場次,參加中央研究院推動、由越南社會科學翰林院主辦的 PNC(the Pacific Neighborhood Consortium)國際學術交流活動,發表多篇關於 AI 輔助佛典解讀、數位化流程與研究應用的實務經驗與研究成果。
透過此次國際會議的交流,CBETA 不僅檢視自身在 AI 導入佛典工作的實際進展,也與來自不同國家與領域的研究者交換經驗,進一步確認人工智慧已成為當代佛教研究與數位人文發展的重要議題之一。
以下將就 CBETA 於 2025 年的主要工作成果,向各位法友簡要說明。
一、佛典資料持續累積,成果清楚可見
截至 2025 年底,CBETA 電子佛典集成累積收錄佛典 4,868 部(21,955 卷),文字總量超過 2.2 億字。僅 2025 年內,即新增 22 部佛典(141 卷)(清單詳見附件一),讓資料庫在內容深度與廣度上持續成長。
這些數字背後,代表的是長時間的掃描、校對、標記與整理工作,確保經文能以穩定且可長期使用的形式保存下來。
此外,2025 年 6 月,CBETA 取得《演培法師全集》之數位化授權,並隨即啟動數位化作業,相關成果亦已陸續完成並公開。演培法師是太虛大師的重要弟子,代表戰後漢傳佛教思想與教育傳承的一支脈絡。其著作內容廣涉經律論講解與弘法文稿,其著作的數位化與收錄,為研究二十世紀佛教發展提供了關鍵而可靠的基礎資料,也為 CBETA 典藏內容注入新的時代面向。
二、經文新式標點持續推進,提升閱讀與教學可用性
CBETA 持續推動佛典的新式標點工作,讓經文在不改變原文內容的前提下,更符合現代讀者的閱讀習慣,也更適合教學與研究使用。
今年在《大正藏》、《嘉興藏》與《卍續藏》等典籍中,皆有實際新增與擴充的新式標點成果(見附件二),使經文段落結構更清楚,語義理解更為順暢。
三、閱藏系統與官網更新,改善實際使用體驗
CBETA 持續優化閱藏系統。單機 Mac 版 CBReader 新增更彈性的全文檢索方式,允許以書目或前次搜尋結果為基礎進行查詢,並支援指定至單一典籍的搜尋範圍。
Windows 版則新增了方便好用的工具箱,不僅可連結諸多線上字辭典、佛經目錄資源、大正藏圖檔,也可以更容易使用各家 AI 平台以及學術研究工具等,這些皆能讓我們在同一個畫面讀經並查詢資料,不僅減少干擾,更能提升效率!
此外,CBETA 官網已於 2025 年 6 月完成改版,資訊架構更為清楚,並支援多語系擴充,讓不同背景的使用者更容易找到所需資源。
四、人工智慧技術導入,強化佛典處理與檢索效能
人工智慧是 2025 年 CBETA 工作中的重要新元素。我們不僅在規劃與研究層面討論 AI,而是已將其實際應用於多個作業環節,除了一般 AI 協助功能之外,包括 OCR 辨識、新式標點,以及語意搜尋工具的測試與營運。
9 月時我們邀請了中研院史語所至 CBETA 舉辦 OCR 工作坊,透過與研究機構合作及內部教育訓練,強化同仁在 OCR 與相關數位技術上的實務理解,使新技術的導入能與既有的學術判斷與品質控管相互配合。
在新式標點等工作流程中,CBETA 採用「AI 產生初稿、人工審核定稿」的協作模式。AI 用於提升整體處理效率,人工作業則負責語意判斷與版本取捨,確保自動化的導入不會犧牲經文品質與學術可靠性。
此外,CBETA Online 的 AI 語意搜尋(Beta 版)已於 2025 年上線。雖仍屬實驗性質,但已累積實際使用經驗,使我們得以逐步評估其效能、成本與適用情境,作為後續優化與發展的重要依據。
五、推廣與交流,讓成果被看見與使用
2025 年,CBETA 透過演講、線上教學與影音內容,持續介紹數位佛典的使用方式與研究可能性;同時也透過社群平台,穩定與使用者互動,使更多人實際接觸並使用 CBETA 的資料與工具。
此外,CBETA 亦逐步將典籍收錄規劃相關資訊對外公開,於官網公布待收錄典籍清單,讓關心佛典收錄進展的讀者與研究者能夠即時查詢,提升資訊透明度與交流互動。
六、資料庫規模與結構的具體進展
以 2025 年度統計來看,CBETA 在典籍部數、卷數與字數上皆有明確成長;同時,CBETA Online 的使用者數已累積至 189 萬人,使用國家數達 198 國,顯示資料庫已被廣泛應用於不同地區與研究場域。
七、校訂考證資料庫持續累積,奠定學術引用的信任基礎
校訂考證資料庫是 CBETA 長期耕耘、卻不易被一般讀者直接察覺的重要基礎建設。凡經文中出現用字修訂、版本差異或必要調整,皆會留下清楚的校訂與考證紀錄,讓使用者能追溯修改的理由與依據。
截至 2025 年底,校訂考證資料庫已累積收錄 48,304 筆 記錄,並與經文資料緊密連結,使研究者在引用 CBETA 文本時,不僅能使用修訂後的經文,也能清楚掌握其形成過程。這些看似細微卻嚴謹的工作,正是 CBETA 能夠被學界長期信賴與引用的關鍵所在。
新的一年,我們將持續以同樣踏實的步伐,維護資料品質,並審慎推進新技術的應用。
誠摯感謝各位法友的長期支持,並邀請您持續關注與使用 CBETA。
謹祝 新春平安吉祥!
釋惠敏 敬上
財團法人佛教電子佛典基金會董事長
2026.1.1
Dear Friends in Dharma, Happy New Year!
Thanks to your support and encouragement, CBETA is about to embark on its 28th year of digitizing the Buddhist canon. Looking back on 2025, we continued to advance the digitization and long-term preservation of Buddhist texts with steady and solid steps. Building upon our existing foundation, we gradually accumulated concrete results that are verifiable and usable.
In 2025, we have expanded the content of our Buddhist text database while deepening foundational projects such as text collation, revision, and modern punctuation. Simultaneously, we have cautiously integrated new technologies into our operational workflows, aiming to enhance overall processing efficiency and user experience without compromising textual quality or scholarly reliability.
Building upon these efforts, CBETA also brought relevant research findings into international academic exchange forums in 2025. In October of that year, CBETA organized a paper session titled “Reimagining Buddhist Scholarship with Artificial Intelligence” at the PNC (Pacific Neighborhood Consortium) international academic conference, hosted by Academia Sinica and the Vietnam Academy of Social Sciences (Viện Hàn lâm Khoa học xã hội Việt Nam). Multiple papers were presented, sharing practical experiences and research outcomes on AI-assisted Buddhist scripture interpretation, digitization workflows, and research applications. Through this international conference, CBETA not only reviewed its practical progress in integrating AI into Buddhist text projects but also exchanged experiences with researchers from diverse countries and disciplines. This further confirmed that artificial intelligence has become a pivotal issue in contemporary Buddhist studies and digital humanities development.
Below is a brief overview of CBETA’s major achievements in 2025 for our Friends in Dharma.
I. Continued Accumulation of Buddhist Canon Data with Tangible Results
By the end of 2025, the CBETA Electronic Buddhist Canon had accumulated 4,868 Buddhist texts (21,955 fascicles), totaling over 220 million characters. In 2025 alone, 22 new Buddhist texts (141 fascicles) were added (see Appendix 1 for details), further expanding the database’s depth and breadth. Behind these figures lies extensive scanning, proofreading, annotation, and organization work to ensure the texts are preserved in a stable and long-term usable format.
Furthermore, in June 2025, CBETA obtained digitalization authorization for The Complete Works of Venerable Yen Pei and immediately commenced digitization efforts. Related outcomes have been progressively completed and made publicly available. Venerable Yen Pei was a key disciple of Master Taixu, representing a significant lineage in post-war Chinese Buddhist thought and education. His writings extensively cover commentaries on sutras, vinaya texts, and treatises, as well as Dharma propagation manuscripts. The digitization and inclusion of his works provide crucial and reliable foundational materials for researching 20th-century Buddhist development.
II. Advancing Modern Punctuation in Sutras to Enhance Readability and Teaching Applications
CBETA continues advancing modern punctuation for Buddhist texts, adapting scriptures to contemporary reading habits without altering original content, thereby enhancing their suitability for teaching and research. This year, tangible progress in adding and expanding modern punctuation has been made in texts such as the Taishō Tripiṭaka, Jiaxing Canon, and Manji Zokuzōkyō (see Appendix II). This work clarifies the structural organization of scripture passages and facilitates smoother semantic comprehension.
III. Reader System and Website Updates for Enhanced Usability
CBETA continues optimizing the Reader system. The standalone Mac version of CBReader now features more flexible full-text search capabilities, allowing queries based on catalog entries or previous search results, and supporting search-within-text functionality for specific texts. The Windows version now features a convenient Toolbox. This tool not only connects to numerous online dictionaries, Buddhist scripture catalogs, and Taishō Canon image archives but also facilitates easier access to various AI platforms and academic research tools. Additionally, the CBETA official website underwent a major redesign in June 2025, featuring a clearer information architecture and multilingual support.
IV. Integration of Artificial Intelligence Technology to Enhance Buddhist Text Processing and Retrieval Efficiency
Artificial intelligence represents a significant new element in CBETA’s 2025 initiatives. We have already implemented it across multiple operational phases, including OCR recognition, modern punctuation, and the testing and operation of semantic search tools. In September, we invited the Academia Sinica Institute of History and Philology to host an OCR workshop at CBETA. For workflows like modern punctuation, CBETA employs a collaborative model where “AI generates drafts, and humans review and finalize”. Additionally, CBETA Online’s AI-powered semantic search (CBETA Semantic Search, CBSS) (Beta version) launched in 2025.
V. Outreach and Exchange, Making Achievements Visible and Usable
In 2025, CBETA continued to introduce the usage and research potential of the Digital Buddhist Canon through lectures, online tutorials, and audiovisual content. CBETA also progressively disclosed information regarding its canon collection planning, publishing a list of texts awaiting inclusion on its official website and social media platforms.
VI. Specific Progress in Database Scale and Structure
Based on 2025 statistics, CBETA has achieved significant growth in the number of texts, fascicles, and characters. Concurrently, CBETA Online has accumulated 1.89 million users across 198 countries, demonstrating the database’s widespread application.
VII. Continuous Accumulation of Revision and Collation Databases, Establishing a Trustworthy Foundation for Academic Citation
The Textual Revision and Verification Database represents a crucial long-term effort by CBETA. Any textual revisions, textual variant adjustments, or necessary modifications are documented with clear records, enabling users to trace the rationale for changes. By the end of 2025, this database had accumulated 48,304 entries, tightly linked to the textual data.
In the new year, we will continue with the same committed pace to maintain data quality and prudently advance the application of new technologies. We sincerely thank all Friends in Dharma for your enduring support and invite you to continue following and utilizing CBETA.
Wishing you peace and blessings in the New Year!
Respectfully,
Bhikshu Huimin (Guo Minfang)
Chairman, CBETA Foundation
January 1, 2026
附件一:CBETA 2025 年新收錄典籍清單
Appendix I: List of Newly Included Texts in 2025
2025.R1
► CBETA 選集 / CBETA Selected Collection(共 1 部 1 卷)
- CC006n0006 佛說大乘無量壽莊嚴清淨平等覺經會集(1卷)【民國 夏蓮居集】
2025.R2
► 佛教大藏經 / Fojiao Canon(共 3 部 70 卷)
- G114-118n2302 (跨冊) 佛學名相彙解(68卷)【明 沙門寂照編】
- G157n2600 始終心要義記(1卷)【民國 倓虛述】
- G160n2619 宗喀巴大師傳(1卷)【民國 張建木述】
2025.R3
► 演培法師全集 / The Complete Works of Venerable Yen Pei(共 13 部 51 卷)
- YP00na001 演培長老傳略(1卷)【演培法師全集出版委員會】
- YP00na002 演培法師全集.出版緣起(1卷)【演培法師全集出版委員會】
- YP00na003 演培法師全集.序文(1卷)【演培法師全集出版委員會】
- YP01n0001 金剛般若波羅密經講記(4卷)【民國 釋演培著】
- YP01n0002 金剛經概要(1卷)【民國 釋演培著】
- YP01n0003 金光明經空品講記(4卷)【民國 釋演培著】
- YP02n0004 般若波羅密多心經講記(3卷)【民國 釋演培著】
- YP02n0005 般若波羅密多心經講記(10卷)【民國 釋演培著】
- YP02n0006 心經十二講(12卷)【日本高神覺昇著 民國 釋演培.關凱圖合譯】
- YP03n0007 佛說十善業道經講記(4卷)【民國 釋演培著】
- YP03n0008 佛說八大人覺經講記(1卷)【民國 釋演培著】
- YP03n0009 大寶積妙慧童女經講記(4卷)【民國 釋演培著】
- YP04n0010 勝鬘師子吼一乘大方便方廣經講記(5卷)【民國 釋演培著】
► 佛教大藏經 / Fojiao Canon(共 3 部 70 卷)
- G151n2559 十二門論講話(1卷)【民國 慈航述】
G155n2589 法寶壇經解義(2卷)【民國 許聖可著】 - G155n2590 六祖大師法寶壇經箋註(13卷)【民國 丁福保註】
- G157n2594 修習止觀坐禪法要講述(1卷)【民國 寶靜述】
- G157n2601 天台傳佛心印記註釋要(2卷)【民國 倓虛釋要】
附件二:2025 年新增新式標點 & AI 自動標點清單
Appendix II: List of Newly Added Modern and AI-Assisted Punctuation in 2025
2025.R1
► 新增新式標點(人工校讀)共 64 部 374 卷
- 大正藏(7 部 59 卷)
- 嘉興藏(56 部 314 卷)
- 卍續藏(1 部 1 卷)
► 新增 AI自動標點共 187 部 490 卷
2025.R2
► 新增新式標點(人工校讀)共 108 部 274 卷
- 嘉興藏 (56 部 183 卷)
- 卍續藏 (52 部 91 卷)
► 新增AI自動標點共 575 部 1903 卷
2025.R3
► 新增新式標點(人工校讀)共 108 部413 卷
- 卍續藏 (108 部 413 卷)
► 新增AI自動標點共 215 部 1341 卷
詳細清單請參考 2025 R1, R2, R3 更新記錄。
For a detailed list, please refer to the 2025 R1, R2, and R3 Update Records.