v2.1.0

• Rebuilt System Settings with clearer grouped controls, an About panel,
  deployment health visibility, local data export/reset actions, and refreshed
  localized settings copy.
• Added native image generation and image editing for models with image
  input/output metadata, including ordered mixed text/image output blocks,
  image edit attachments, and OPFS-backed display caching.
• Expanded built-in plugin media tools: Agnes and Gemini now present as image
  processing plugins, OpenAI-compatible Images API and OpenAI Responses image
  processing are separate built-ins, and image plugin results are compacted into
  tool details/history so follow-up model messages decide how to reference them.
• Added plugin-level API Base URL and Model ID controls for supported image
  plugins, image count parameters where the upstream API supports them, Agnes
  image-to-image editing, and Agnes video image-to-video support with custom
  video model IDs while preserving the two-step create_video /
  get_video_result workflow.
• Added thinking intensity controls and provider-specific reasoning mapping for
  Gemini and OpenAI-compatible model requests.
• Added Japanese localization across the app, SEO metadata, LobeHub assistant
  locale routing, voice language handling, and the public Skills catalog.
• Hardened hosted deployments with API request proof, stronger shared-store and
  rate-limit checks, service health coverage, safer URL/secret handling, and
  expanded test coverage.
• Fixed Cloudflare Workers preview/deploy commands and kept Worker deploys from
  dropping dashboard-managed variables.
• Refined code block rendering, syntax highlighting, sandboxed HTML preview,
  Mermaid/mind map/SVG rendering behavior, and release automation based on
  matching CHANGELOG.md sections.
• Added a fork-only upstream sync workflow and README guidance for keeping fork
  repositories current with u14app/neo-chat.

v2.0.0

v2.0.0 is a completely new build. Please see README.md for version details.

You can find previous versions on the gemini-next-chat branch, which is archived and no longer accepts any modification requests.

• Added open-source governance files, issue templates, pull request template,
  Dependabot configuration, and documentation for environment variables,
  plugin development, and privacy/data handling.
• Added required Prettier format checking to CI after a one-time repository
  formatting pass.
• Added text-only Skills with localized public catalogs, install/uninstall,
  local edits, custom skills, auto-selection, and workspace presets.
• Expanded message rendering with safe inline HTML visual blocks, Mermaid and
  mind map fullscreen rendering, richer source blocks, and visible search
  failure states.
• Hardened hosted and multi-instance deployment behavior with shared plugin
  registry storage, document parse job secrets, deployment health checks,
  trusted proxy guidance, and safer sandbox/document parsing limits.
• Added local memory documentation and Mimo voice defaults alongside existing
  search, RAG, document parsing, and BYOK configuration guidance.

v1.10.1

feat: The latest model list will be automatically pulled when the page is loaded
fix: Fix the problem of invalid regeneration
fix: Fixed the issue of abnormal layout of search references
fix: Models that don't support plugins no longer send tools settings

---

feat: 页面加载时会自动拉取最新的模型列表
fix: 修复重新生成功能无效问题
fix: 修复搜索引用布局异常问题
fix: 不支持插件的模型不再发送工具设置

v1.10.0

This update focuses on enhanced image and text capabilities and performance optimizations:

• New Image and Text Model: Added support for the gemini-2.0-flash-exp-image-generation model, which can generate content with mixed image and text layouts.
• Image Size Optimization: Generated images will be automatically compressed to reduce file size.
• Image Loading Optimization: Optimized image loading using lazy loading technology to reduce the impact of image rendering on text generation speed.

Now, you can experience richer image and text interaction and smoother content generation speed!

---

本次更新聚焦于更强大的图文能力和性能优化：

• 全新图文模型: 新增支持 gemini-2.0-flash-exp-image-generation 模型，可生成包含图文混排的内容。
• 图片体积优化: 生成的图片将自动压缩，减小文件大小。
• 图片加载优化: 采用懒加载技术优化图片加载，减少图片渲染对文本生成速度的影响。

现在，您可以体验更丰富的图文互动和更流畅的内容生成速度！

v1.9.1

fixed: the issue that the client could not use real-time multimodal video and screen capture normally, and optimized the settings interface appropriately.

---

修复了客户端无法正常使用实时多模态视频和屏幕捕捉的问题，并适当优化了设置界面。

v1.9.0

This update introduces powerful multimodal live streaming capabilities and optimizes performance and documentation:

• Major Release: Multimodal Live Streaming!
  • Supports the Gemini Multimodal Live API. Note: Currently, the official version only supports models like gemini-2.0-flash.
• Voice Mode Upgrade: The original voice mode has been refactored into a component, and automatic recording settings have been added.
• Performance Optimization:
  • Optimized Office file parsing code, changed to dynamic import, reducing the number of files loaded on the landing page.
  • The system instruction and attachment area components are now dynamically loaded, improving loading speed.
  • Removed early compatibility code from the store.
• Documentation Update:
  • Improved documentation content and added a new Roadmap.
  • Added Frequently Asked Questions for the Multimodal Live API.
  • Added documentation for using the Multimodal Live API proxied by Cloudflare Worker.
• Bug Fixes:
  • Adjusted defaultValue to value to prevent form state from being cached.
• Build Adjustments: Adjusted the wrangler.toml configuration.

We are committed to continuously optimizing the product experience. Stay tuned for more exciting features!

---

本次更新带来了强大的多模态直播功能，并对性能和文档进行了优化：

• 重磅推出：多模态直播！
  • 支持 Gemini Multimodal Live API。注意：目前官方仅支持 gemini-2.0-flash 这类模型。
• 语音模式升级： 将原语音模式重构为组件，并增加自动录音设置。
• 性能优化：
  • 优化了Office文件解析代码，改为动态导入，减少落地页加载的文件数量。
  • 将系统指令和附件区域组件改为动态加载，提升加载速度。
  • 移除了 store 中的早期兼容代码。
• 文档更新：
  • 改进了文档内容，并添加了新的路线图（Roadmap）。
  • 新增了多模态直播 API 常见问题解答。
  • 新增了使用 Cloudflare Worker 代理的多模态直播 API 文档。
• Bug修复：
  • 调整 defaultValue 为 value，防止表单状态被缓存。
• 构建调整： 调整了 wrangler.toml 配置。

我们致力于不断优化产品体验，敬请期待更多精彩功能！

v1.8.1

This update primarily focuses on user interface enhancements and model optimization:

• Model switching and title display in the conversation interface: You can now switch models within the conversation interface, and the conversation title is displayed for easier management and use.
• File list display optimization: File names in the file list are now displayed more clearly and understandably.
• Removed outdated models: Cleaned up unused models to maintain a concise model list.
• CSS style adjustments: Optimized the interface style to enhance the visual experience.

---

本次更新主要集中在用户界面和模型优化：

• 会话界面模型切换与标题显示: 现在可以在会话界面切换模型，并显示会话标题，方便管理和使用。
• 文件列表显示优化: 文件列表中的文件名显示更加清晰易懂。
• 移除过时模型: 清理了不再使用的模型，保持模型列表的简洁。
• CSS 样式调整: 优化了界面样式，提升视觉体验。

v1.8.0

This update focuses on enhancing file processing capabilities and conversation experience:

• New Office File Parsing: You can now parse the content of Office files (e.g., .docx, .xlsx, .pptx).
• Office File Upload Support: The file upload feature now supports Office file types.
• Smarter Conversation Management: Empty conversations are no longer automatically named.
• Optimized Issue Output Experience: Improved fluency and readability of problem presentation.
• Bug Fixes:
  • Fixed an issue where Cloudflare pages did not support fetch cache.
  • Fixed an issue where text-type files could not be uploaded correctly.
  • Fixed an issue with abnormal functionCall variable judgment.

---

本次更新重点提升了文件处理能力和对话体验：

• 新增 Office 文件解析: 现在可以解析 Office 文件（例如 .docx, .xlsx, .pptx）的内容。
• 支持上传 Office 文件: 文件上传功能支持 Office 文件类型。
• 更智能的对话管理: 空对话不再自动命名。
• 优化问题输出体验: 提升了问题呈现的流畅性和易读性。
• Bug 修复:
  • 修复了 Cloudflare 页面不支持 fetch cache 的问题。
  • 修复了文本类型文件无法正常上传的问题。
  • 修复了 functionCall 变量判断异常的问题。

v1.7.2

This update mainly enhances the Google Grounding Search function:

• New Google Grounding Search result UI: You can now view the search results of Google Grounding Search more intuitively.
• Document export adaptation to Grounding Search: The document export function has been optimized to better support Grounding Search.
• Added error information to the model list API: The model list API now returns more detailed error information to facilitate troubleshooting.
• Document update: Corrected the incorrect document description and added more detailed instructions.

---

本次更新主要围绕 Google Grounding Search 功能进行了增强：

• 新增 Google Grounding Search 结果UI: 现在可以更直观地查看 Google Grounding Search 的搜索结果。
• 文档导出适配 Grounding Search: 优化了文档导出功能，使其更好地支持 Grounding Search。
• 模型列表API增加错误信息: 模型列表API现在会返回更详细的错误信息，方便问题排查。
• 文档更新: 更正了错误文档描述，并添加了更详细的说明。

v1.7.1

Support for Gemini 2.0

---

支持 Gemini 2.0