v2.0.17

2026-06-16 · v2.0.17

中文

• 新增跨会话添加页面：编辑时可从其他已有会话中选择页面复制到当前演示末尾，适合复用历史内容、合并方案或快速拼接多份稿件。
• 新增主会话页面范围选择：对整套演示发起 AI 修改时，可选择全部页面或指定页面范围，避免一次改动影响不相关页面。
• 新增批量编辑进度面板：整套/多页修改会实时展示每页处理状态、失败原因和完成结果，失败页面可单独重试。
• 新增整套风格切换：可在编辑工作台中选择新风格并重绘整套演示，尽量保留原有文字、数据和页面含义，同时重新设计视觉布局。
• 新增编辑标尺、网格与辅助线：编辑画布支持吸附、参考线和标尺定位，拖拽排版、对齐元素和统一页面结构更方便。
• 新增会话搜索：会话列表支持按标题搜索，历史演示较多时更容易定位目标会话。
• 新增会话与模板封面缩略图生成：会话列表、模板卡片和风格预览的封面加载更完整，并支持缩略图缺失时自动补全。
• 新增 Token 用量统计：设置中可查看模型调用次数、输入/输出 Token、按天趋势、当天小时趋势和不同模型的用量分布（仅供参考）。
• 新增 OpenAI Responses Provider：基础模型配置可选择 OpenAI Responses 接口，兼容新的 OpenAI 调用链路。
• 新增 Linux 打包配置，可自行打包
• 新增风格包导入导出：支持导入风格 ZIP 或风格文件夹，也可将单个风格导出为 ZIP；风格页新增场景筛选、删除确认和官方风格解析 Skill 入口。
• 增强PPTX 导入：进一步改进表格、图表和页面结构解析，导入后的页面更容易继续编辑、生成缩略图和复用为风格。
• 重构：风格系统内置风格从 30+ 扩展到 70+，风格资源改为独立文件夹结构，风格生成skill用：官方技能包[https://github.com/arcsin1/style-generate-skill]。
• 优化 OpenAI 兼容模型参数处理：官方 OpenAI 端点不再携带非标准 thinking 参数，其他兼容端点仍会按需关闭 thinking，减少 400 Unknown parameter 和推理字段兼容问题。
• 优化所有页面的ui风格：所有页面都ui增强，支持预览缩略图以及交互增强。
• 优化页面生成与修改摘要：生成完成说明会基于已验证页面数据构建，编辑摘要会隔离模型原始输出，减少完成提示与实际页面状态不一致的问题。
• 优化会话列表 UI 与长列表渲染：页面缩略图按需加载，列表滚动和大量会话展示更顺畅，性能增强。

English

• Added pages from another session: select slides from an existing session and copy them to the end of the current deck, making it easier to reuse previous work, merge proposals, or assemble decks quickly.
• Added page range selection for main-session AI edits: apply deck-level instructions to all slides or only selected slides, reducing accidental changes to unrelated pages.
• Added a batch edit progress panel: track the status, failure reason, and final result for each slide during whole-deck or multi-slide edits, with retry support for failed slides.
• Added full-deck style switching: choose a new style in the editor and redraw the whole presentation while keeping the original text, data, and slide meaning as much as possible.
• Added editor rulers, grids, and guides: snapping and reference lines make alignment, layout, and consistent slide structure easier.
• Added session search: search session titles from the session list to find older decks faster.
• Added cover thumbnail generation for sessions and templates: session lists, template cards, and style previews now show covers more completely, and missing thumbnails can be filled automatically.
• Added token usage statistics: view model call counts, input/output tokens, daily trends, today's hourly usage, and usage distribution by model from Settings.
• Added OpenAI Responses Provider: base model configuration can now choose the OpenAI Responses interface.
• Added Linux packaging configuration, so Linux builds can be packaged manually.
• Improved OpenAI-compatible model handling: official OpenAI endpoints no longer receive unsupported thinking parameters, while compatible endpoints still request thinking to be disabled when needed, reducing related configuration and call failures.
• Reworked the style system: built-in styles expanded from 30+ to 70+, with a new style resource structure. Use the official style generation skill at [https://github.com/arcsin1/style-generate-skill].
• Added style package import/export: import style ZIPs or style folders, export individual styles as ZIP files, filter styles by use case, confirm deletion, and open the official style parser skill from the Styles page.
• Improved the UI style across all pages: every page has enhanced visuals, preview thumbnails, and interaction details.
• Enhanced PPTX import: tables, charts, and slide structure are parsed more accurately, making imported slides easier to edit, thumbnail, and reuse as styles.
• Improved generation and edit summaries: completion messages are now based on confirmed slide results, and edit summaries avoid exposing raw model output, reducing mismatch between messages and actual slide state.
• Improved the Sessions page and long-list performance: slide thumbnails load as needed, scrolling through many sessions is smoother, and overall performance is better.

v2.0.16

中文

• 新增 MP4 视频导出功能：可将整套演示导出为视频文件，适合分享、归档或在不方便播放 PPT 的场景中直接使用。
• 新增页面切换动画：支持淡入、滑动、推进、擦除、缩放、翻转、立方体、封面流、模糊、聚焦展开等多种切页效果，演示节奏更接近正式放映（只在浏览器打开和演示模式有效）。
• 新增鼠标滚轮翻页：演示和浏览时可使用鼠标滚轮切换页面，翻页手感更自然，并减少连续滚动导致的误触。
• 新增导出进度提示：PPTX、图片和视频导出过程中会显示实时进度、当前阶段和完成结果，长文件导出时更安心。
• 增强 PPTX 导出质量：导出的可编辑 PPTX 对文字、图片、图表、表格、动画和页面结构的还原更稳定，复杂页面在 PowerPoint 和 WPS 中的表现更可靠（因为没有大量测试模版，不能做到100%还原，大概80%-95%效果）。
• 增强 PPTX 导出速度：长演示和复杂页面的导出等待时间减少，批量导出体验更顺畅。
• 增强 PPTX 导入：导入已有 PPTX 时可更好识别图表、表格和页面内容，导入后的页面更容易继续编辑和复用。
• 优化页面画布：生成页与预览页减少默认边距干扰，页面更接近真正的全屏幻灯片效果，背景和主体内容的边界更一致。
• 优化生成布局质量：AI 会先根据内容密度选择合适的版式，再进行宽高自检，减少内容超出画布、上下堆叠和卡片过密的问题，同时保留更自由的创意布局。
• 优化图表页面生成：图表页会更明确地区分主图表、辅助信息和注释，减少图表过高、信息区挤压或多行卡片堆叠的问题。
• 优化大纲与资料生成：当用户明确列出多个要点时，页面规划会更完整地保留这些重点，减少漏点、合并不当或结构偏移。
• 优化生成完成提醒：生成任务通知更清晰，完成、失败和部分完成状态更容易识别。
• 修复思考文档刷新异常：部分场景下思考内容不及时更新的问题已改善。

English

• Added MP4 video export: export an entire presentation as a video for sharing, archiving, or playback when a PPT file is not the best fit.
• Added slide transitions: choose from fade, slide, push, wipe, zoom, flip, cube, cover flow, blur, iris, swing, and center reveal for a more presentation-ready rhythm.
• Added mouse wheel navigation: move between slides with the mouse wheel while browsing or presenting, with smoother control and fewer accidental jumps.
• Added export progress feedback: PPTX, image, and video exports now show live progress, the current stage, and the final result, making long exports easier to follow.
• Improved PPTX export quality: editable PPTX files now preserve text, images, charts, tables, animations, and slide structure more reliably, with better results in PowerPoint and WPS.
• Improved PPTX export speed: long presentations and complex slides now export with less waiting, making batch export smoother.
• Improved PPTX import: existing PPTX files now restore charts, tables, and slide content more accurately, so imported slides are easier to edit and reuse.
• Improved the slide canvas: generated and previewed slides now feel closer to true full-screen slides, with less default spacing around the canvas.
• Improved generated layout quality: AI now chooses layouts based on content density and checks width and height before writing, reducing overflow, vertical stacking, and overcrowded cards while keeping layouts more creative.
• Improved chart slide generation: chart pages now better separate the main chart, supporting information, and notes, reducing oversized charts, squeezed content, and stacked support cards.
• Improved outline and source-based generation: when users provide explicit points, slide planning preserves them more completely and reduces missing points, unwanted merging, or structural drift.
• Improved generation completion notifications: task updates are clearer, making completed, failed, and partially completed states easier to recognize.
• Fixed thinking document refresh issues in some sessions.

v2.0.15

2026-06-07 · v2.0.15

中文

• 新增多任务生成：创建演示会话后可继续发起其他任务，无需停留在生成页面等待；会话列表会持续展示排队、生成进度和完成状态。
• 新增应用内任务通知：生成完成、部分完成或失败时会在应用内及时提醒，并可直接进入对应会话查看结果。
• 新增会话另存为：可复制当前演示的页面与素材创建独立新会话，适合保留原稿后继续制作不同版本。
• 新增浏览模式：在编辑工作台中以更纯净的方式连续查看整套演示，编辑、浏览和演示等使用场景切换更清晰。
• 新增艺术字：编辑页面可直接插入艺术字，并可以调整文案、效果和字号，标题与重点内容的视觉表现更丰富。
• 新增模型兼容性设置：基础模型配置支持关闭 temperature 参数，兼容不接受该参数的模型服务。
• 新增对话创作大纲编辑：可在生成前手动修改每页的标题、页面目标、内容摘要和关键要点，后续对话与生成会使用更新后的大纲。
• 大幅增强资料理解与大纲解析：上传 Markdown、文本、CSV、DOCX 或图片后，可更准确地提炼章节、关键结论、数据和页面结构，让生成内容更贴近原始资料。
• 增强对话创作：AI 能更持续地整理上传资料、创作目标和页面规划，在生成前形成更清晰、可调整的内容方案。
• 增强 PPTX 导入：支持更好地还原表格、图表、页面大纲和动画信息，复杂演示导入后的可编辑性与结构完整度进一步提升。
• 优化多模型任务执行：不同任务可分别选择运行模型，支持使用多个模型处理生成、编辑、模板复用、文档解析和演讲稿等任务。
• 优化编辑工作台：重新组织顶部工具、页面导航、AI、生图、属性、演讲稿和历史记录等功能，常用操作更集中，编辑状态更清晰。
• 优化编辑历史：撤销、重做和页面编辑记录更稳定，复杂编辑过程中的回退与恢复更可靠。
• 优化大文件与长演示支持：演示页数上限提升至 500 页，PPTX 上传大小上限提升至 500MB。
• 优化缩略图加载：长演示的页面缩略图按需加载，进入编辑页和滚动页面列表时更流畅。
• 优化模板创建与复用：文档解析建议、模型选择和生成参数确认更集中，使用模板创建新内容时更直观。
• 修复 macOS Apple Silicon 环境下部分会话包无法正常打开的问题。

En

• Added multi-task generation: start another presentation while existing tasks continue in the background, with queued, generating, and completed states visible from the Sessions page.
• Added in-app task notifications: receive a notification when generation completes, partially completes, or fails, then open the related session directly.
• Added Save as New Session: duplicate the current slides and assets into an independent session, making it easier to preserve the original while creating variations.
• Added Browse mode: review the full presentation in a cleaner workspace, with clearer transitions between editing, browsing, and presenting.
• Added WordArt: insert stylized text directly in the editor and adjust its content, effect, and size for more expressive titles and highlights.
• Added a model compatibility option: Base Model Configuration can omit the temperature parameter for services that do not support it.
• Added outline editing in Chat to Create: revise each slide's title, objective, summary, and key points before generation; later conversations and generation use the updated outline.
• Significantly improved source understanding and outline parsing: Markdown, text, CSV, DOCX, and image uploads are now better organized into chapters, key findings, data points, and slide structures.
• Enhanced Chat to Create: AI now maintains a clearer understanding of uploaded sources, creative goals, and slide plans before turning them into a generation-ready outline.
• Enhanced PPTX import: tables, charts, slide outlines, and animation metadata are restored more accurately for better editability and structural fidelity.
• Improved multi-model task execution: different tasks can use different models for generation, editing, template reuse, document parsing, speaker scripts, and more.
• Improved the editing workspace: reorganized tools for page navigation, AI, image generation, properties, speaker scripts, and history so common actions are easier to find and current modes are clearer.
• Improved edit history: undo, redo, and slide edit records are more reliable during longer or more complex editing sessions.
• Improved support for large files and long presentations: presentations can now contain up to 500 slides, and PPTX uploads can be up to 500MB.
• Improved thumbnail loading: slide thumbnails are loaded on demand for smoother entry and scrolling in long presentations.
• Improved template creation and reuse: document suggestions, model selection, and generation settings are now presented in a more focused workflow.
• Fixed an issue that could prevent some session packages from opening on Apple Silicon Macs.

v2.0.14

这是一个功能性增加和优化的过渡版本，计划下个版本对编辑页进行大的交互改造！
Starting from v2.0.14, this project is licensed under Apache License 2.0. Earlier releases remain under their original license terms.

中文

• 新增 AI 生图工作流：编辑页右侧新增图片生成模式，可围绕当前页面生成配图素材，并支持将生成图片添加到画布、设置为页面背景或直接定位本地文件。
• 新增多生图模型接入配置：目前支持即梦、Agnes AI、硅基流动、OpenAI-compatible 图片接口和 Gemini 图片生成，并可在设置页独立管理、启用和验证生图模型。
• 新增配图提示词辅助：支持「大纲提示词」和「风格提示词」，可根据当前页标题、大纲、页面视觉风格和用户输入生成更适合当前页面的配图描述；
• 新增统一模型切换体验：创建、模板复用、风格导入、图片解析等需要模型参与的操作，都可以在执行前选择本次使用的模型，减少频繁进入设置页切换的成本，多模型切换更方便。
• 新增页面大纲：编辑页支持读取、展示和编辑修改页面大纲，创建、导入、页面管理和单页补写流程中的页面上下文更清晰。
• 新增 HTTP/SOCKS 代理支持：可为模型服务配置代理，提升访问 Gemini 等外部服务时的网络兼容性。
• 新增 macOS Intel 架构支持配置，提升 Apple Silicon 与 Intel Mac 的安装包兼容性。
• 新增添加文字功能：编辑页可直接添加文字元素，文字编辑体验进一步优化。
• 优化对话创作： 增加对话历史记录，用户可以查看和管理之前的对话记录，方便后续参考和继续创作。
• 优化设置页：增加一些配置帮助说明。
• 优化模板复用体验：从模板创建演示时的参数确认、模型选择和生成前校验更完整，减少模板结构异常导致的失败。
• 优化导入与解析稳定性：文档解析、大纲归一化、PPTX 导入和风格抽取流程更稳，对复杂输入的容错更好。
• 优化编辑体验：改进文字选中、字体编辑、图层检查和元素锚点定位，减少重叠元素或复杂页面中的误选和编辑偏差。
• 优化技能提示词：更新图表与布局技能说明，生成图表、版式和页面结构时更贴合演示文稿场景。

En

• Added AI image generation: create slide visuals from the editor, then add them to the canvas, set them as the slide background, or reveal the generated file locally.
• Added image model configuration: connect Jimeng, Agnes AI, SiliconFlow, OpenAI-compatible image APIs, and Gemini image generation from Settings.
• Added image prompt helpers: use "Outline prompt" and "Style prompt" to turn the current slide outline, visual style, and your own idea into a better image description.
• Added model selection at run time: choose which model to use before creation, template reuse, style import, image parsing, and other AI-powered actions.
• Added slide outlines: the editor can now read, display, and edit slide outlines, making the current slide context clearer during editing, import, page management, and single-slide regeneration.
• Added HTTP/SOCKS proxy support: configure a proxy for model services to improve connectivity with external providers such as Gemini.
• Added macOS Intel support: packaging configuration now better supports both Apple Silicon and Intel Macs.
• Improved chat-to-create: conversation history is now easier to keep, review, and continue from during creative planning.
• Improved Settings: added helpful configuration guidance.
• Improved template reuse: stronger parameter confirmation, model selection, and pre-generation checks reduce failures from invalid template structure.
• Improved import and parsing stability: document parsing, outline normalization, PPTX import, and style extraction now handle complex inputs more reliably.
• Added text insertion: add text elements directly in the editor, with further improvements to text editing.
• Improved editing: text selection, font editing, layer inspection, and element anchoring are more accurate on complex or overlapping slide elements.
• Improved built-in skill prompts: chart and layout instructions now better match presentation-generation scenarios.

十分感谢大家支持！目前主要由我一人开发和维护，如果您觉得 Oh My PPT 对您有所帮助，可以赞助支持一下：

支持赞助token

v2.0.13

中文

• 新增动画效果增强：生成页面时支持更丰富的入场动画效果，包括淡入、飞入、擦除、缩放、旋转等，并支持逐条出现和交错展示，演示内容可按讲述节奏逐步展开。
• 新增 PPTX 动画导出：导出可编辑 PPTX 时保留动画效果和页面切换转场，导出后可在 PowerPoint或者wps中直接演示播放。
• 新增 PPTX 动画导入：导入 PPTX 时自动解析并还原原稿的动画效果，无需手动重新设置。
• 新增富文本编辑：编辑文字时可使用加粗、斜体、下划线、颜色和字号等格式，编辑体验更接近专业文档工具。
• 优化元素选中精度：点击重叠或嵌套元素时能更精准地定位到目标文字，不再被上层元素遮挡或误选。
• 优化生成质量：页面布局、图表和动画的生成效果更稳定，版式更合理、图表更准确、动画更自然。
• 优化更新提醒：新版本通知改为弹窗，可查看版本对比、选择下载方式和查看更新日志。
• 优化创建防重复提交：创建演示稿时防止重复点击导致创建多个会话。

English

• Added animation enhancements: generated slides now support richer entrance animations including fade, fly-in, wipe, zoom, and spin, with staggered reveals and click-by-click display so content unfolds with the speaker's rhythm.
• Added PPTX animation export: exported editable PPTX files now preserve animation effects and slide transitions, playable directly in PowerPoint.
• Added PPTX animation import: imported PPTX files now automatically parse and restore the original animation effects, no manual setup needed.
• Added rich text editing: edit text with bold, italic, underline, color, and font size formatting for a more professional editing experience.
• Improved element selection precision: clicking overlapping or nested elements now accurately targets the intended text, no longer blocked or misdirected by upper-layer elements.
• Improved generation quality: slide layout, charts, and animations are more stable and accurate with better composition and more natural motion.
• Improved update notification: new version alerts now appear as a dialog with version comparison, download options, and changelog link.
• Improved session creation guard: prevents duplicate sessions from accidental double-clicks.

v2.0.12

中文

• 新增模板库：支持将已生成或已编辑的演示保存为模板，管理模板名称、描述和标签，并在模板页统一预览、编辑、删除和复用。
• 新增从模板创建ppt会话：可直接复用模板生成可编辑会话，也可输入新主题/大纲或上传文档解析后，沿用模板视觉系统重新生成内容。
• 新增 PPTX 文件可导入为模板：模板页可导入 .pptx 作为模板，自动解析页面、抽取视觉风格并生成模板设计契约。
• 新增页面标题编辑：编辑页支持修改单页标题。
• 新增空白页创建：可选择已有页面作为版式基础创建到末尾，并自动清空可见文字，适合快速延展同风格页面。
• 优化首页 PPTX 快速导入编辑入口：首页可直接导入 .pptx 创建可编辑会话，并展示导入进度和提示信息。
• 优化编辑/检选体验：进入编辑和 AI 检选时会结束残留动画，避免带动画页面出现元素隐藏、停在半透明状态或无法选中的问题。
• 优化视频元素编辑与播放：可以对视频元素进行编辑，包括调整播放速度、音量、循环播放、自动播放等。
• 优化导入与运行时稳定性：会话导入更好地兼容扁平 ZIP 和复杂压缩包，导入会话补齐默认设计契约；
• 修复图表问题：图表类目轴bug、横向柱状图 tooltip 和 0 值数据展示更准确。
• 修复页面管理问题：修复页面管理时的动画问题，避免页面切换时元素隐藏、停在半透明状态或无法选中的问题。

English

• Added Template Library: save generated or edited presentations as templates, manage template names, descriptions, and tags, then preview, edit, delete, and reuse them from the Templates page.
• Added PPT session creation from templates: reuse a template to create an editable session directly, or enter a new topic/outline or upload a document to regenerate content while keeping the template's visual system.
• Added PPTX import as templates: import .pptx files from the Templates page, automatically parse slides, extract the visual style, and create a template design contract.
• Added page title editing: edit individual slide titles from the editor.
• Added blank page creation: choose an existing slide as the layout base, create a new page at the end of the deck, and clear visible text for fast same-style page expansion.
• Improved the quick PPTX import editing entry on Home: import a .pptx directly from the Home page to create an editable session with progress and warning feedback.
• Improved edit and inspect modes: residual animations are finished before editing or AI inspection, reducing hidden, half-transparent, or hard-to-select elements on animated slides.
• Improved video element editing and playback: video elements can now be edited, including playback speed, volume, loop playback, autoplay, and more.
• Improved import and runtime stability: session import better handles flat ZIPs and complex archives, and imported sessions receive a default design contract.
• Fixed chart issues: category-axis bugs, horizontal bar chart tooltips, and zero-value data now render more accurately.
• Fixed page management issues: resolved animation problems during page management, reducing hidden, half-transparent, or hard-to-select elements when switching pages.

v2.0.11

2026-05-24 · v2.0.11

中文

• 新增对话创作模式：先通过多轮对话梳理主题、资料、受众、结构和每页重点，再确认参数生成完整演示稿。适合需求尚不清晰、资料较复杂或需要先共同推敲大纲的场景，再去生成创作（轻智能体可能不太稳定）。
• 新增演讲稿创作：支持为整套幻灯片或当前页生成演讲稿，内置正式演讲、轻松对话、叙事风格和自定义风格，并可复制或直接查看文件。 (感谢： @whisper-xiang )
• 新增 Gemini Provider：设置页支持配置 Google Gemini，可用于生成、编辑和对话创作。
• 重构动画系统：支持更自然的入场动画、错峰展示和点击逐条出现，让演示内容可以按讲述节奏逐步展开。 (感谢：@Jacobinwwey)
• 优化演示播放：键盘导航、鼠标点击和逐条展示的衔接更顺畅，带点击节奏的页面不容易误翻页。
• 优化预览与导出稳定性：带动画、图表和复杂内容的页面在预览、缩略图和导出时显示更可靠。
• 优化生成预览： 生成的时候支持边生成边预览，无需等待生成完成，方便确认生成效果。

English

• Added Chat to Create: first use a multi-turn conversation to clarify the topic, materials, audience, structure, and key points for each slide, then confirm the parameters and generate the full presentation. It is designed for unclear requirements, complex materials, or scenarios where you want to shape the outline together before creation.
• Added speaker script creation: generate scripts for the full deck or the current slide, with built-in formal, casual conversational, storytelling, and custom styles. Scripts can be copied or opened directly as a file.
• Added Gemini Provider: Google Gemini can now be configured in Settings and used for generation, editing, and Chat to Create.
• Rebuilt the animation system: supports more natural entrance animations, staggered reveals, and click-by-click content display, so presentation content can unfold with the speaker's rhythm.
• Improved presentation playback: keyboard navigation, mouse clicks, and step-by-step reveals now work together more smoothly, making click-paced pages less likely to accidentally advance to the next slide.
• Improved preview and export stability: slides with animations, charts, and complex content now render more reliably in previews, thumbnails, and exports.
• Improved generation preview: slides can now be previewed while they are still being generated, so you no longer need to wait for the full generation to finish before checking the result.

Other

Many thanks for the contributions below！

@whisper-xiang @Jacobinwwey

v2.0.10

新增

• 新增元素属性面板：选中元素后可独立调整文字样式（字号、粗细、颜色）、外观（背景色、边框、圆角、阴影）、布局（尺寸、位置）、图层（层级、可见性）和媒体属性，编辑更精细。
• 新增取色器：颜色选择支持自定义取色板和透明度调节，编辑颜色更灵活直观。
• 新增导出会话文件：编辑页面一键导出会话生成的创意 PPT，另一台电脑导入后可继续二次编辑，跨设备协作无缝衔接。
• 新增导入会话文件：会话列表支持导入 slide-pack 打包的文件或标准会话 ZIP，导入后可继续二次编辑。
• 新增编辑页面历史记录：编辑模式下自动记录操作历史，支持在保存版本前浏览和回退到任意编辑步骤。
• 新增字体数据回填：升级后自动为已有会话补充字体信息，兼容历史数据。

优化

• 优化生成成功率：大幅提升页面生成的稳定性和成功率，减少生成失败和格式异常，增加兜底策略，减少 token 消耗。
• 优化生成引擎：改进大纲规划与页面写入逻辑，生成结果更完整、布局更合理。
• 优化编辑交互：元素选中和拖拽体验更流畅，编辑模式整体更稳定。

修复

• 修复风格预览 bug：兼容 win 下的风格预览问题，展示更直观。

---

New

• Element Properties Panel: After selecting an element, you can independently adjust text styles (font size, weight, color), appearance (background color, border, corner radius, shadow), layout (size, position), layers (z-index, visibility), and media properties for more precise editing.
• Color Picker: Supports custom color palettes and transparency adjustment, making color editing more flexible and intuitive.
• Export Session File: One-click export of creative PPT sessions from the editing page. The file can be imported on another computer for further editing, enabling seamless cross-device collaboration.
• Import Session File: The session list supports importing slide-pack packaged files or standard session ZIP files. After import, you can continue editing.
• Edit History: The editing mode automatically records operation history, allowing you to browse and roll back to any editing step before saving the version.
• Font Data Backfill: Automatically supplements font information for existing sessions after upgrade, ensuring compatibility with historical data.

Optimized

• Generation Success Rate: Significantly improves the stability and success rate of page generation, reduces generation failures and formatting errors, adds fallback strategies, and reduces token consumption.
• Generation Engine: Improves outline planning and page writing logic, resulting in more complete outputs and more reasonable layouts.
• Editing Interaction: Enhances the smoothness of element selection and dragging, making the overall editing mode more stable.

Fixed

• Style Preview Bug: Fixed style preview issues on Windows for a more intuitive display.

v2.0.9

V2.0.9

尽管现阶段已支持导出PPTX格式，但最终的愿景是：以AI驱动的HTML PPTX格式替代传统PPTX，实现在浏览器内直接演示

中文

• 新增字体管理：内置 14 款精选 Google 字体（含中文），支持上传本地 .woff2 字体，可自定义字体名称、分类、用途和语言类型。
• 新增字体选择：创建演示稿时可分别指定标题字体和正文字体，也可交给 AI 根据主题和风格自动匹配。
• 新增字体嵌渲染实质化：导出 PPTX 时自动嵌入已使用的字体文件，确保在 PowerPoint 中打开后字体不丢失、不回退。
• 新增风格缩略图：风格列表展示可视化预览缩略图，挑选风格更直观。
• 新增模型 Max Tokens 设置：可在设置中自定义模型最大输出长度，适配不同模型的响应上限。
• 新增 Windows 系统托盘：关闭窗口后应用最小化到系统托盘，后台运行不干扰桌面。
• 优化风格管理：整体风格编辑和展示体验更流畅，显示风格的适用场景。
• 优化内置风格： 重写内置风格的提示词和style_skill。
  - 优化 PPTX 导出：导出流程和效果持续改进（暂不支持视频导出以及动画导出，html的复杂度导致导出无法完全一样）。
• 优化文字大小：优化生成默认最小文字为16px，更适合ppt风格的预览和演示。
• 修复：演示模式和本地预览的键盘鼠标bug
• 修复：Windows 11 下批量导出图片失败的问题。
• 修复：修复默认导出目录不正确的问题。

En

• Added font management: 14 curated Google Fonts built-in (including CJK), with support for uploading local .woff2 fonts and customizing name, category, role, and script.
• Added font selection: choose title and body fonts separately before generation, or let AI auto-match based on topic and style.
• Added font embedding in PPTX export: used fonts are automatically embedded in the exported file so they display correctly in PowerPoint without fallback.
• Added style thumbnails: the style list now shows visual preview thumbnails for easier selection.
• Added Max Tokens setting: customize the model's maximum output length in Settings to match different model limits.
• Added Windows system tray: closing the window minimizes the app to the system tray instead of quitting.
• Improved style management: smoother editing and browsing experience, with applicable scenarios shown for each style.
• Improved built-in styles: rewrote prompts and style skills for all built-in styles.
• Improved PPTX export: further refinements to the export pipeline and output quality.
• Improved minimum font size: default minimum text size is now 16px for better readability in previews and presentations.
• Fixed keyboard and mouse issues in presentation mode and local preview.
• Fixed batch PNG export failing on Windows 11.
• Fixed incorrect default export directory.

十分感谢大家支持！目前主要由我一人开发和维护，如果您觉得 Oh My PPT 对您有所帮助，可以赞助支持一下：

支持赞助token

v2.0.8

🎉🎉🎉🎉 V2.0.8

增加一个独特的功能： 一键打包当前的html pptx为单个可执行文件（类似 PPTX），随时随地双击即可打开预览，无需安装任何软件（有浏览器就行）
- Added one-click HTML pack: bundle the current HTML presentation into a single executable file — double-click to open and present anywhere, no installation needed (just a browser).

中文

• 优化数学公式导出：可编辑版 PPTX 中，公式以截图形式作为独立图片插入，确保在 PowerPoint 中正确显示。
• 优化pptx导出，支持更多元素
• 新增： 一键打包当前的html pptx为单个可执行文件（类似 PPTX），随时随地双击即可打开预览，无需安装任何软件（你有浏览器就行）。

English

• Improved math formula export: formulas are captured as individual images in editable PPTX for correct display in PowerPoint.
• Improved PPTX export functionality with support for additional elements.
• Added one-click HTML pack: bundle the current HTML presentation into a single executable file — double-click to open and present anywhere, no installation needed (just a browser).