v1.3.9

English | 中文

Highlights

• Added official mpt pip packaging. Start the WebUI with mpt webui or generate videos with mpt cli.
• Added static and slow zoom image effects for local and AI-generated images. Slow zoom remains the default.
• Added Kimi Code subscription access with browser authorization, alongside improvements to the sign-in dialog, status updates, and cancellation.
• Added an optional local SDXL image server for the OpenAI-compatible image material workflow, with idle GPU offloading.
• Added Opper, onomeo, Muse Spark, and EmpirioLabs AI LLM providers, and completed Brazilian Portuguese WebUI translations.
• Added diagnostic logging controls in the WebUI, including runtime log-level changes and opt-in API request tracing with redaction and size limits.

Fixes and Improvements

• Fixed grayscale image brightness, transparent image backgrounds, zoom mask alignment, and subtitle background colors. Improved transitions on short clips, background music fade-out, and special-character handling in concatenation paths.
• Isolated intermediate image files across concurrent jobs, improved image cache identity, and validated cached downloads. Completed videos are retained when another video in the same batch fails.
• Fixed speech speed, volume, and audio container handling across several providers. Improved remote audio validation and preserved existing audio until replacement output is ready.
• Improved subtitle timestamp rounding, long-word wrapping, Unicode handling, and pause-tag recognition. Symbol-only lines and separators no longer distort estimated subtitle alignment.
• Fixed Chinese and Japanese fallback social titles using the entire script. Improved JSON extraction, reasoning removal, Markdown cleanup, and generated search-term validation.
• Expanded supported local upload formats, including multi-picture camera JPEGs. Improved task history checks, deletion state, subtitle previews, and model connection error messages in the WebUI.
• Added HEAD support for generated artifacts and improved byte-range downloads, validation errors, and Redis task updates and deletion safeguards.
• Reduced unsafe speech resubmissions after lost responses or redirects, preserved Upload-Post tracking handles, and improved WaveSpeed submission error detection.
• Updated Edge TTS to address 403 failures, constrained PyAV for Whisper compatibility, and fixed Windows cache cleanup and UTF-8 BOM handling in locale files.
• Improved the agent helper's configuration handling and preflight checks, including workflows with prepared narration and visuals that do not need an LLM.

What's Changed

• Official pip package with WebUI and CLI launchers — c3634de.
• Static and slow zoom image options — 90f7701.
• Kimi Code subscription provider — @ysntony in #1599, followed by sign-in flow improvements.
• Local SDXL image server — @imkashif in #1668.
• Opper LLM provider — @Felixkw12 in #1589.
• onomeo LLM provider — @Owaid0723666 in #1600.
• Muse Spark provider and Brazilian Portuguese translations — @humbertocortezia in #1657.
• EmpirioLabs AI provider — @Adam-Dalloul in #1694.
• Runtime diagnostic logging and request tracing — 6940611.
• Chinese and Japanese fallback social titles — e25326e.
• Retained completed videos after batch failures — @hdu-r3drum in #1597.
• Edge TTS 403 fix — @HEOJUNFO in #1652.
• Whisper/PyAV compatibility — @cristianchiriac in #1582.
• Windows cache cleanup — @shan3520 in #1654.
• UTF-8 BOM locale support — @ketpatil77 in #1658.
• Task protection for unparseable states — @RerankerGuo in #1663.
• Clearer model connection errors — @fengqiaoqu in #1645.
• Local material preflight checks — @drakeo338 in #1585; preserved Redis task IDs — @ac1982 in #1588.
• Grayscale rendering, artifact HEAD requests, and prepared-media agent workflows — @rudycelekli in #1620, #1636, and #1608, alongside many subtitle, speech, material, API, and task reliability fixes.

Thanks to all contributors and everyone who reported issues and helped test these changes!

Full Changelog: v1.3.8...v1.3.9

---

重点更新

• 支持通过 pip 安装官方 mpt 包，使用 mpt webui 启动界面，或通过 mpt cli 生成视频。
• 新增图片动态效果选项，本地图片和 AI 生成图片均可选择“静止”或“缓慢放大”，默认保留缓慢放大。
• 新增 Kimi Code 订阅接入，支持浏览器授权登录，并优化登录窗口布局、状态更新和取消操作。
• 新增可选的本地 SDXL 生图服务，通过 OpenAI 兼容接口接入图片素材生成，支持空闲时释放 GPU 显存。
• 新增 Opper、onomeo、Muse Spark 和 EmpirioLabs AI 大模型服务，完善巴西葡萄牙语界面翻译。
• 新增诊断日志设置，支持在 WebUI 中即时调整日志等级，按需记录经过脱敏和长度限制的 API 请求参数。

修复与改进

• 修复灰度图片亮度、透明图片背景、缩放蒙版对齐和字幕背景颜色问题；改善短片段转场、背景音乐淡出和拼接路径中的特殊字符处理。
• 隔离并发任务的图片中间文件，改善图片缓存标识并校验下载缓存；批量生成部分失败时，保留已经完成的视频。
• 修复部分配音服务的语速、音量和音频格式处理，完善远程音频校验；替换音频准备完成前保留已有文件。
• 改善字幕时间戳舍入、长词换行、Unicode 字符和停顿标签识别，避免纯符号行和分隔线干扰字幕时间轴估算。
• 修复中文、日文备用社交标题使用整篇脚本的问题；改善 JSON 提取、推理内容过滤、Markdown 文案清理和素材搜索词校验。
• 补齐支持的本地上传格式，包括多图相机 JPEG；改善 WebUI 历史视频检查、任务删除状态、字幕预览和模型连接错误提示。
• 支持生成文件的 HEAD 请求，完善范围下载、参数错误提示，以及 Redis 任务状态更新和删除保护。
• 减少语音请求在响应丢失或重定向后被重复提交的风险，保留 Upload-Post 请求追踪信息，并改善 WaveSpeed 提交失败的识别。
• 更新 Edge TTS，修复相关 403 问题；限制 PyAV 版本以兼容 Whisper，修复 Windows 缓存清理和带 UTF-8 BOM 的翻译文件读取。
• 改善 Agent 辅助脚本的配置处理和启动检查，已准备好配音与画面的任务不再额外要求大模型服务。

主要变更

• c3634de：官方 pip 安装包与 WebUI、CLI 启动命令。
• 90f7701：图片静止与缓慢放大选项。
• #1599：Kimi Code 订阅接入，感谢 @ysntony；后续补充了登录流程优化。
• #1668：本地 SDXL 生图服务，感谢 @imkashif。
• #1589：Opper 大模型服务，感谢 @Felixkw12。
• #1600：onomeo 大模型服务，感谢 @Owaid0723666。
• #1657：Muse Spark 大模型服务与巴西葡萄牙语翻译，感谢 @humbertocortezia。
• #1694：EmpirioLabs AI 大模型服务，感谢 @Adam-Dalloul。
• 6940611：运行时诊断日志设置与请求追踪。
• e25326e：中文、日文备用社交标题修复。
• #1597：批量生成失败后保留已完成的视频，感谢 @hdu-r3drum。
• #1652：Edge TTS 403 修复，感谢 @HEOJUNFO。
• #1582：Whisper 与 PyAV 兼容性修复，感谢 @cristianchiriac。
• #1654：Windows 缓存清理修复，感谢 @shan3520。
• #1658：带 UTF-8 BOM 的翻译文件支持，感谢 @ketpatil77。
• #1663：状态无法解析时的任务保护，感谢 @RerankerGuo。
• #1645：更清晰的模型连接错误提示，感谢 @fengqiaoqu。
• #1585：本地素材启动检查，感谢 @drakeo338；#1588：保留 Redis 任务 ID 原始字符串，感谢 @ac1982。
• #1620、#1636、#1608：灰度图片渲染、生成文件 HEAD 请求和已备素材的 Agent 工作流，以及多项字幕、配音、素材、API 与任务稳定性修复，感谢 @rudycelekli。

感谢所有贡献代码、反馈问题和参与测试的朋友！

完整变更记录：v1.3.8...v1.3.9

v1.3.8

English | 中文

Highlights

• Added continuous progress reporting during stock footage downloads, clip processing, and final rendering, with task logs from background helper threads visible in the WebUI.
• Added configurable stock material and clip rendering concurrency. Stock material concurrency applies to Pexels, Pixabay, and Coverr; both settings default to 1 (serial processing).
• Added VoxCPM reference-audio voice cloning and optional delivery conditioning to carry over pacing and emotion, with local Whisper transcription for editable reference transcripts.
• Added Fluxion AI, Cheaper Inference, Requesty, FutureInfra, Y-API, and iFlytek Spark (Astron MaaS) LLM providers.
• Added MuAPI as a video material source, with model endpoint configuration and asynchronous generation.
• Added Catalan WebUI localization.
• Improved video color consistency by using the BT.709 conversion matrix and color metadata in the main rendering pipeline.
• Added an opt-in local video project CLI with revision tracking, change comparison, and reuse of unchanged rendering artifacts when replacing scenes, narration, or background music.

Fixes and Improvements

• Improved Redis task updates, queue admission, and startup recovery to reduce lost fields, scheduling conflicts, and stalled queues. Invalid queued requests no longer prevent later usable tasks from being dispatched.
• Added deadlines for stalled FFmpeg operations and improved temporary-file cleanup. Failed encodes no longer replace successful video outputs with incomplete files.
• Improved concurrent rendering by isolating concatenation manifests and intermediate clips across jobs and video combinations.
• Streamed remote video downloads into the cache, added download limits and validation, and improved cache identity and cleanup to avoid reusing the wrong asset or removing fresh downloads.
• Improved local material handling, including supported upload formats, image decoding, camera orientation, CMYK colors, and rejection of album artwork mistakenly identified as video footage.
• Preserved timestamps appearing in subtitle text, punctuation in Whisper word results, and manual line breaks. Disabled subtitles no longer pick up stale caption files.
• Improved subtitle timing fallback and concurrent Whisper initialization. Empty subtitle results now fail explicitly instead of publishing caption files without cues.
• Hardened narration generation across multiple providers: validate audio before replacing existing files, preserve successful narration when export fails, and stop unsafe retries after synthesis has already been accepted.
• Improved pause-segment assembly and Azure timing conversion, and fixed narration-only exports ignoring the configured audio volume.
• Improved script and keyword generation validation, preventing provider error messages from being accepted as generated content. Updated Groq’s retired default model and passed Claude Code prompts through stdin.
• Added WebUI preflight validation for local materials and custom narration. Improved settings-cache serialization, active task refresh, historical-data handling, version checks, and failed audio-upload cleanup.
• Hardened uploaded filenames, subtitle font paths, media size limits, and render parameters. Expanded Windows smoke coverage and fixed subprocess output decoding to use UTF-8.
• Rejected redirects for several credentialed or potentially paid requests, reducing credential exposure and accidental repeat submissions.
• Improved Upload-Post result validation and recovery tracking. Queued publishing retains the selected account, and YouTube privacy settings are preserved even without generated metadata.
• Improved API download URLs, special-character escaping, media types, range handling, boolean parsing, CORS origin normalization, and validation-error responses.
• Updated provider documentation, example configuration, and the agent helper to better match supported CLI features and material sources.

Usage and Upgrade Notes

• Both concurrency settings default to 1. Higher stock material concurrency may trigger provider rate limits; higher clip rendering concurrency uses more CPU and memory. Increase them according to your environment.
• VoxCPM reference audio and delivery transcripts are sent to ModelBest when synthesis is requested. Use only your own voice or recordings you are authorized to use. Optional local Whisper transcription may download the configured model on first use.
• The local video project workflow is CLI-only and initially scoped to prepared local footage and narration. It does not add a WebUI video editor or automatically generate scripts, speech, subtitles, or transitions. Retained source snapshots and revisions require additional disk space.
• Custom subtitle fonts must resolve inside the project’s resource/fonts directory. External paths and symlinks pointing outside that directory are rejected.
• Providers or gateways that redirect generation requests now require their final API endpoint to be configured directly. An unconfirmed paid request stops local processing rather than being automatically resubmitted; this does not imply that the initial request was not charged.
• New provider integrations require the corresponding API credentials and service configuration. iFlytek Spark’s pay-as-you-go and Token Plan endpoints use different credentials; select the endpoint matching your account.
• The existing WebUI generation entry point remains unchanged. New concurrency options, reference-audio controls, and the local project CLI can be used as needed.

What's Changed

• Generation progress reporting and background task logs — @Amitk2108 in #1518.
• Configurable material and clip rendering concurrency — @keepliu28 in #1407, followed by serial defaults and clearer WebUI controls.
• VoxCPM reference voice cloning and delivery conditioning — @lottshin in #1390.
• MuAPI video material source — @Anil-matcha in #1367.
• Fluxion AI LLM provider — b66674b.
• Cheaper Inference LLM provider — @aiapienthusiast in #1401.
• Requesty LLM provider — @Thibaultjaigu in #1473.
• FutureInfra LLM provider — @uxidev in #1524.
• Y-API LLM provider — @jiweiyeah in #1530.
• iFlytek Spark (Astron MaaS) LLM provider — @FenjuFu in #1531.
• Catalan WebUI localization — fdcf249.
• BT.709 video encoding and color metadata — @HEOJUNFO in #1527.
• Revision-aware local video projects — @rudycelekli in #1549.
• CLI support for zoom transitions, subtitle display options, and WaveSpeed — @c020627 in #1357, #1362, and #1368, alongside further CLI and agent-helper fixes.
• UTF-8 subprocess output decoding — @britorj in #1365.
• Timestamp preservation inside subtitle text — @haimuhaimu in #1391.
• Pickle-safe WebUI cache statistics — @thinkw in #1397.
• Updated Groq default model — @RooMax in 26edbbd.
• Claude Code prompt delivery through stdin — @TanPham2808 in #1435.
• Atomic Redis task updates and persisted queue recovery — @rudycelekli in #1409 and #1414, alongside extensive task, media, and publishing reliability fixes.
• Safer paid audio retries, preserved YouTube privacy, and stopped image submission redirects — @rudycelekli in #1516, #1550, and #1551.

Thanks to all contributors and everyone who reported issues and helped test these changes!

Full Changelog: v1.3.7...v1.3.8

---

重点更新

• 新增素材下载、片段处理和最终渲染的持续进度反馈，WebUI 同时支持显示后台辅助线程产生的任务日志，便于了解生成过程是否仍在运行。
• 新增库存素材并发和片段渲染并发设置。库存素材并发适用于 Pexels、Pixabay 和 Coverr，两项默认均为 1（串行处理）。
• 新增 VoxCPM 参考音频音色复刻与演绎参考，支持延续节奏和情绪，并可使用本地 Whisper 生成可编辑的示范音频逐字稿。
• 新增 Fluxion AI、Cheaper Inference、Requesty、FutureInfra、Y-API 和讯飞星辰 MaaS 大模型服务。
• 新增 MuAPI 视频素材来源，支持配置模型接口和异步视频生成。
• 新增加泰罗尼亚语 WebUI 翻译。
• 视频主渲染流程统一使用 BT.709 色彩转换与标记，改善不同播放器中的颜色一致性。
• 新增按需启用的本地视频项目 CLI，支持修订记录、变更比较和渲染结果复用；替换场景、配音或背景音乐时，可减少未变化部分的重复渲染。

修复与改进

• 改进 Redis 任务状态更新、队列准入和启动恢复，减少字段丢失、调度竞争和排队停滞；无法执行的排队请求不再阻塞后续可用任务。
• 为长时间停滞的 FFmpeg 操作增加超时，并完善临时文件清理；编码失败时不再用不完整文件覆盖已成功生成的视频。
• 隔离并发任务和不同视频组合的拼接清单与中间片段，减少渲染过程中的文件冲突。
• 素材下载改为流式写入缓存，增加大小限制和有效性校验，并改进缓存标识与清理逻辑，避免误用素材或删除刚更新的缓存。
• 完善本地素材处理，包括上传格式、图片解码、手机照片方向...

v1.3.7

English | 中文

Highlights

• Added word-by-word subtitles and a pop-up spring animation, with independent controls for display mode and animation in the WebUI.
• Added preset background music selection with in-browser audio preview, so you can listen before generating a video.
• Added native pauses to scripts with Azure TTS V1 (Edge TTS): use [pause: 2s] or [停顿: 2s] to insert silence while keeping narration and subtitles aligned.
• Added Kokoro and VoxCPM TTS providers. Kokoro supports a self-hosted OpenAI-compatible service and automatically loads the server's voice list.
• Added Claude Code as an LLM provider using an authenticated Claude subscription, with dedicated Docker deployment files; also added the OpenAI-compatible API Route provider.
• Improved Shengsuan Cloud video generation with account-specific model discovery, quotes and explicit cost confirmation, plus clip-count recommendations based on narration duration.
• Reduced repeated material selection across videos generated in the same task when using Pexels, Pixabay, Coverr, or local files. Material reuse is reported when the available pool is insufficient.
• Added a YouTube “Made for kids” audience setting to Upload-Post publishing.
• Added Azerbaijani WebUI localization and updated translations for the new settings and prompts.

Fixes and Improvements

• Improved subtitle animation rendering, including synchronized scaling of subtitle images and transparency masks to avoid black outlines when subtitles appear.
• Hardened Kokoro voice discovery and audio handling: retain the selected voice when the service is unavailable and validate generated audio before replacing existing files.
• Added pause-tag validation and improved multi-segment narration/subtitle synchronization; invalid speech chunks now stop generation rather than silently producing incomplete narration.
• Handled non-finite voice rates such as NaN and infinity more safely.
• Fixed the WebUI launch script's port check when the project path contains spaces, and corrected imports and repository paths in the Colab notebook.
• Added periodic log messages while FFmpeg concatenation is running, making long-running processing easier to diagnose.
• Extended material cache cleanup to reclaim temporary files left behind by interrupted writes.
• Updated the agent helper to support openai_image and recognize claude_code as a keyless LLM provider.
• Synced the documented Edge TTS voice list with the built-in catalog and clarified where to find the Windows one-click package in release Assets.

Usage and Upgrade Notes

• Subtitle defaults remain sentence display with no animation. Word-by-word timing works best with Azure TTS V1 (Edge TTS) or Whisper; providers without word-level timestamps may display phrases or whole sentences instead.
• Script pause tags are supported only with Azure TTS V1 (Edge TTS), with individual pauses from 0.1 to 10 seconds.
• Kokoro requires a separately running compatible TTS service. Claude Code requires its CLI and subscription authentication; selecting either provider does not install or authenticate the underlying service automatically.
• The YouTube audience setting is an explicit user declaration, not automatic content classification. It defaults to not made for kids; review the setting before uploading.
• Batch material allocation reduces reuse within a single task; it does not guarantee unique footage when there are too few source clips and does not change paid AI material providers' allocation behavior.

What's Changed

• Word-by-word subtitles and spring animation — @farhanahmad15 in #1313.
• Preset background music selection — @farhanahmad15 in #1314, followed by audio preview improvements.
• Claude Code subscription provider — @fred-ng in #1319.
• WebUI launcher support for paths containing spaces — @tailieu2212 in #1321.
• Kokoro TTS provider — @rocketlang in #1325.
• Shengsuan Cloud video workflow improvements — adapted from @huanglvjing's contribution in #1328, incorporated in 5ceffd0.
• Native script pauses — @IsmaelCamC in #1330.
• Non-finite voice rate handling — @Sushanth012 in #1333.
• Colab notebook fixes — @FrioNode in #1334.
• Azerbaijani WebUI translation — @jamalkamaladdin in #1339.
• Reduced batch material reuse — @giteePro in #1340.
• Edge TTS voice catalog documentation — @dajiaohuang in #1341.
• VoxCPM TTS provider — @lottshin in #1343.
• Orphaned material cache temporary-file cleanup — @c020627 in #1345.
• Release Assets download guidance — @c020627 in #1349.
• FFmpeg concatenation progress logging — @c020627 in #1351.
• Agent helper support for image generation and keyless Claude Code — @c020627 in #1352 and #1353.
• API Route LLM provider — @DennyHo0917 in #1354, with README updates by @c020627 in #1355.
• YouTube audience declaration — bfd7612.

Thanks to all contributors and everyone who reported issues and helped test these changes!

Full Changelog: v1.3.6...v1.3.7

---

重点更新

• 新增逐词字幕和弹跳出现动画，WebUI 可分别选择字幕显示模式与动画效果。
• 新增预设背景音乐选择与浏览器内试听，生成视频前即可确认配乐效果。
• 新增 Azure TTS V1（Edge TTS）脚本停顿支持：使用 [停顿: 2s] 或 [pause: 2s] 插入静音，并同步调整配音与字幕时间轴。
• 新增 Kokoro 和 VoxCPM 配音服务。Kokoro 支持自托管的 OpenAI 兼容服务，并自动读取服务端音色列表。
• 新增通过已认证 Claude 订阅使用的 Claude Code 大模型服务，提供专用 Docker 部署文件；同时新增 OpenAI 兼容的 API Route 大模型服务。
• 优化胜算云视频生成流程，支持读取当前账号可用模型、获取报价和确认费用，并根据旁白时长推荐素材数量。
• 减少同一任务批量生成多个视频时的素材重复，适用于 Pexels、Pixabay、Coverr 和本地文件；素材不足而发生复用时会给出提示。
• Upload-Post 新增 YouTube“是否面向儿童”受众设置。
• 新增阿塞拜疆语 WebUI 翻译，并补齐新设置与提示的多语言文案。

修复与改进

• 优化字幕动画渲染，同步缩放字幕画面与透明蒙版，避免字幕出现时产生黑色轮廓。
• 完善 Kokoro 音色读取与音频处理：服务暂时不可用时保留原有音色选择，替换文件前校验生成音频。
• 增加停顿标签校验，改进分段配音与字幕同步；无效语音片段会中止生成，避免静默输出缺失旁白的成片。
• 完善 NaN、无穷大等非有限语速值的处理。
• 修复项目路径包含空格时 WebUI 启动脚本的端口检查，并修正 Colab notebook 的缺失导入和仓库路径。
• FFmpeg 拼接期间增加定期日志，便于判断长时间处理状态和排查问题。
• 素材缓存清理支持回收写入中断后遗留的临时文件。
• Agent 辅助脚本新增 openai_image 支持，并正确识别无需 API Key 的 claude_code 服务。
• 将 Edge TTS 文档音色列表与内置音色库同步，并明确 Windows 一键包在 Release Assets 中的下载位置。

使用与升级说明

• 字幕默认仍为按句显示、无动画。逐词时间轴在 Azure TTS V1（Edge TTS）或 Whisper 下最准确；其他服务缺少逐词时间戳时，可能按短语或整句显示。
• 脚本停顿标签仅支持 Azure TTS V1（Edge TTS），单次停顿范围为 0.1～10 秒。
• Kokoro 需要单独运行兼容的配音服务；Claude Code 需要安装 CLI 并完成订阅认证。选择服务商不会自动完成底层服务的安装或认证。
• YouTube 受众设置由用户主动声明，不会自动识别视频内容。默认非面向儿童，上传前请根据实际内容确认。
• 批量素材分配仅减少同一任务内的重复；素材不足时仍可能复用，不保证每条视频的画面完全不同，也不改变付费 AI 素材服务的分配行为。

主要变更

• #1313：逐词字幕与弹跳动画，感谢 @farhanahmad15。
• #1314：预设背景音乐选择，感谢 @farhanahmad15；后续补充了音乐试听。
• #1319：Claude Code 订阅大模型服务，感谢 @fred-ng。
• #1321：修复含空格路径下的 WebUI 启动检查，感谢 @tailieu2212。
• #1325：Kokoro 配音服务，感谢 @rocketlang。
• #1328：吸收并优化胜算云视频生成流程，感谢 @huanglvjing，已通过 5ceffd0 纳入。
• #1330：脚本原生停顿支持，感谢 @IsmaelCamC。
• #1333：非有限语速值处理，感谢 @Sushanth012。
• #1334：修复 Colab notebook，感谢 @FrioNode。
• #1339：阿塞拜疆语 WebUI 翻译，感谢 @jamalkamaladdin。
• #1340：减少批量视频素材重复，感谢 @giteePro。
• #1341：同步 Edge TTS 音色文档，感谢 @dajiaohuang。
• #1343：VoxCPM 配音服务，感谢 @lottshin。
• #1345：清理素材缓存遗留临时文件，感谢 @c020627。
• #1349：完善 Release Assets 下载指引，感谢 @c020627。
• #1351：FFmpeg 拼接进度日志，感谢 @c020627。
• #1352、#1353：Agent 辅助脚本支持文生图与免 API Key 的 Claude Code，感谢 @c020627。
• #1354：API Ro...

v1.3.6

English | 中文

Highlights

• Added native Volcano Engine Ark Seedance video generation for generating source clips directly from script segments.
• Added OFox multi-model text-to-video support, allowing Seedance, Wan, and other models to share one API key.
• Added Metaso MiniMax H3 text-to-video support with 768P/2K output, 4–15 second clips, and 9:16, 16:9, and 1:1 aspect ratios.
• Added an OpenAI-compatible text-to-image material source for cloud relays and local gateways.
• Added OpenRouter and APIMart as OpenAI-compatible LLM providers.
• Added a WebUI Auto-Publish settings tab for Upload-Post, with clearer API key and account setup guidance.
• Added CLI batch manifest mode and applied saved WebUI settings as CLI defaults when command-line values are omitted.
• Added configurable source clip fit modes: cover fills the frame by cropping excess edges, while contain preserves the complete source with letterboxing.
• Added complete French and Korean WebUI translations.
• Reorganized material source settings, grouped video sources by capability, and centralized AI video provider credentials in the WebUI settings.

Security

• Restricted browser cross-origin API access to same-origin requests by default. Trusted standalone browser frontends can be enabled explicitly with CORS_ALLOWED_ORIGINS.
• Added provenance and SBOM attestations to published container images.

Fixes and Improvements

• Measured narration duration from the generated audio file instead of relying only on subtitle cues.
• Anchored the SiliconFlow subtitle timeline to the complete audio duration.
• Ensured AudioFileClip resources are closed in ElevenLabs, Chatterbox, and Fish Audio paths.
• Prevented script cleanup from greedily removing unrelated bracketed or parenthesized text.
• Removed a misleading retry warning after the final LLM attempt.
• Fixed Redis URLs when no password is configured.
• Preserved log records when the project root and log destination are on different mounts.
• Fixed Unicode console output after successful CLI generation on Windows.
• Added safe, portable download filenames, including RFC-compatible quoting and protection from Windows-reserved names.
• Improved WebUI preview behavior on headless servers by falling back to browser playback and download.
• Decoupled Upload-Post configuration from automatic publishing so settings can be saved without enabling uploads.
• Hardened OpenAI-compatible image handling: malformed 200 responses now skip only the affected keyword, while local storage failures stop further paid requests.

Upgrade Notes

• Browser API access now defaults to same-origin requests. Standalone web frontends must declare their trusted origins through CORS_ALLOWED_ORIGINS; server-side clients such as curl, Postman, and n8n are unaffected.
• CLI commands now inherit saved WebUI settings for omitted options. Explicit CLI arguments and batch manifest values still take precedence.

What's Changed

• Add Fish Audio to the English and Japanese provider documentation by @c020627 in #1246
• Add container provenance and SBOM attestations by @kobihikri in #1140
• Add the Upload-Post Auto-Publish WebUI settings by @parveen0029 in #1209
• Improve headless WebUI video preview behavior by @abhiunix in #1252
• Apply saved WebUI settings as CLI defaults by @Gainto in #1225
• Decouple Upload-Post configuration from automatic publishing by @mcgeand in #1259
• Preserve logs across different filesystem mounts by @SandroHub013 in #1260
• Add CLI batch manifest mode by @lihuiyang1024 in #1243
• Fix API download filename quoting by @Mihir7027 in #1265
• Use human-readable batch manifest size errors by @Mihir7027 in #1264
• Fix Redis URLs without passwords by @Mihir7027 in #1266
• Remove the final spurious LLM retry warning by @Mihir7027 in #1267
• Anchor SiliconFlow subtitles to the full audio duration by @Mihir7027 in #1268
• Measure narration duration from the real audio file by @YUSAKRU in #1263
• Close TTS audio clips reliably by @Mihir7027 in #1269
• Make script bracket and parenthesis cleanup non-greedy by @Mihir7027 in #1270
• Add APIMart OpenAI-compatible LLM support in ab1c790
• Add OpenRouter LLM support by @brizzio in #1277
• Fix successful CLI runs on consoles that cannot encode Unicode by @housine35 in #1282
• Add configurable source clip fit modes by @emanuelsc3s in #1287
• Avoid Windows-reserved WebUI download filenames by @Sushanth012 in #1288
• Add the OpenAI-compatible text-to-image material source by @joker1point in #1291
• Skip malformed successful image responses without aborting the task by @joker1point in #1296
• Add Korean WebUI localization based on @greekr4's contribution in #1193
• Add French WebUI localization based on @PocketBinary's contribution in #1248
• Add native Volcano Engine Ark Seedance support based on @lihuiyang1024's contribution in #1276
• Add OFox multi-model video generation by @celeste1900 in #1292
• Add Metaso MiniMax H3 video generation and centralized AI video provider settings in 44ce430
• Restrict cross-origin browser access based on the security report from @AkshatJoshi2000.

Full Changelog: v1.3.5...v1.3.6

---

重点更新

• 新增火山引擎方舟 Seedance 原生视频生成服务，可根据脚本片段直接生成视频素材。
• 新增 OFox 多模型文生视频服务，一个 API Key 即可调用 Seedance、Wan 等模型。
• 新增秘塔 MiniMax H3 文生视频服务，支持 768P/2K、4～15 秒片段，以及 9:16、16:9 和 1:1 三种画幅。
• 新增适用于云端中转和本地网关的 OpenAI 兼容文生图素材源。
• 新增 OpenRouter 和 APIMart OpenAI 兼容大语言模型服务商。
• WebUI 新增 Upload-Post 自动发布设置，并完善 API Key 和账号名称的申请、配置指引。
• CLI 新增批量清单模式；未显式传入命令行参数时，可以继承 WebUI 已保存的默认设置。
• 新增素材画面适配模式：cover 通过裁剪边缘填满画面，contain 保留完整素材并在比例不同时增加黑边。
• 新增完整的法语和韩语 WebUI 翻译。
• 重新整理素材来源设置，按能力对视频来源分组，并将 AI 视频服务商凭据统一集中到 WebUI 设置中管理。

安全性

• 浏览器跨域 API 访问现在默认遵循同源策略；如需使用独立网页前端，可以通过 CORS_ALLOWED_ORIGINS 显式配置可信来源。
• 发布的容器镜像新增构建来源证明和 SBOM 软件物料清单证明。

修复与改进

• 改为从真实生成的音频文件测量旁白时长，不再仅依赖字幕时间轴。
• 将 SiliconFlow 字幕时间轴末尾与完整音频时长对齐。
• 确保 ElevenLabs、Chatterbox 和 Fish Audio 链路始终关闭 AudioFileClip 资源。
• 修复脚本文本清理可能贪婪删除多组括号内容的问题。
• 移除大语言模型最后一次重试结束后误导性的“正在重试”日志。
• 修复 Redis 未配置密码时生成错误鉴权片段的问题。
• 修复项目目录和日志目录位于不同挂载点时日志记录丢失的问题。
• 修复 Windows 终端无法编码 Unicode 时，CLI 在视频成功生成后异常退出的问题。
• 改进下载文件名的安全性和跨平台兼容性，包括标准引用编码以及 Windows 保留名称保护。
• 无图形桌面的服务器环境下，WebUI 会回退到浏览器播放和下载视频。
• 将 Upload-Post 配置与自动发布开关解耦，允许用户保存配置但不立即启用上传。
• 加强 OpenAI 兼容文生图异常处理：HTTP 200 返回损坏图片时仅跳过当前关键词，本地存储失败时则停止后续付费请求。

升级说明

• 浏览器调用 API 现在默认遵循同源策略。独立网页前端需要通过 CORS_ALLOWED_ORIGINS 配置可信来源；curl、Postman、n8n 等服务端调用不受影响。
• CLI 未显式传入参数时，会继承 WebUI 保存的设置；明确传入的 CLI 参数和批量清单字段仍拥有更高优先级。

主要变更

• #1246：在英文和日语文档中补充 Fish Audio，感谢 @c020627。
• #1140：为容器镜像增加构建来源证明和 SBOM，感谢 @kobihikri。
• #1209：新增 Upload-Post 自动发布 WebUI 设置，感谢 @parveen0029。
• #1252：改进无图形桌面服务器上的 WebUI 视频预览，感谢 @abhiunix。
• #1225：CLI 支持继承 WebUI 已保存设置，感谢 @Gainto。
• #1259：将 Upload-Post 配置和自动发布解耦，感谢 @mcgeand。
• #1260：修复跨文件系统挂载时日志记录丢失，感谢 @SandroHub013。
• #1243：新增 CLI 批量清单模式，感谢 @lihuiyang1024。
• #1265：修复 API 下载文件名引用编码，感谢 @Mihir7027。
• #1264：使用更易读的批量清单大小错误提示，感谢 @Mihir7027。
• #1266：修复 Redis 未配置密码时的连接 URL，感谢 @Mihir7027。
• #1267：移除最后一次 LLM 尝试后的错误重试提示，感谢 @Mihir7027。
• #1268：将 SiliconFlow 字幕与完整音频时长对齐，感谢 @Mihir7027。
• #1263：从真实音频文件测量旁白时长，感谢 @YUSAKRU。
• #1269：可靠释放 TTS 音频资源，感谢 @Mihir7027。
• #1270：修复脚本括号内容清理的贪婪匹配，感谢 @Mihir7027。
• ab1c790：新增 APIMart OpenAI 兼容大语言模型服务商。
• #1277：新增 OpenRouter 大语言模型服务商，感谢 @brizzio。
• #1282：修复部分 Windows 终端...

v1.3.5

English | 中文

Highlights

• Added MiniMax and Fish Audio text-to-speech providers, improved MiniMax regional endpoint and voice selection support, synchronized Gemini voices with the official catalog, and added automatic voice previews in the WebUI.
• Added Anthropic Claude as a native LLM provider.
• Added Shengsuan AI and WaveSpeed text-to-video material sources.
• Added reusable WebUI generation settings and preset import/export, including optional API key backup and restore.
• Added LLM-generated social captions for every supported cross-post platform, instead of limiting them to YouTube.
• Added an Italian WebUI translation and a Japanese README.

Security

• Added optional API key authentication for /api/v1 endpoints and generated task files. Existing installations remain backward compatible while app.api_key is empty.
• Restricted custom audio inputs to allowed locations to prevent path traversal.
• Prevented generated task files from being accessed through escaping symlinks.
• Hardened local material uploads with stricter file and path validation.
• Sanitized client-supplied request IDs before using them in task processing.

Fixes and Improvements

• Fixed the last subtitle line being clipped for some text and font combinations, including multiline subtitles and text without descenders.
• Added an actionable startup error when FFmpeg is unavailable.
• Fixed Docker dependency installation retries so persistent package failures stop the build instead of being silently ignored.
• Rolled back task state when background scheduling fails, preventing tasks from remaining stuck in an incorrect state.
• Improved API response schemas and migrated model configuration to Pydantic v2 ConfigDict.
• Registered the ping health-check router.
• Preserved paragraph boundaries in normalized LLM responses.
• Distinguished the supported Kimi API platforms in configuration and the WebUI.
• Handled configuration files containing repeated UTF-8 BOM markers.
• Improved MiniMax voice discovery and validated generated audio before replacing an existing output file.

Upgrade Notes

• No authentication is required by default. To protect the V1 API and generated task files, set app.api_key in config.toml and send the same value in the x-api-key request header.
• Settings preset exports can contain provider credentials when API key backup is enabled. Store exported preset files securely and do not commit them to source control.
• Fish Audio, MiniMax, Anthropic Claude, WaveSpeed, and Shengsuan AI require their respective provider credentials before they can be used.

What's Changed

• Add MiniMax text-to-speech support by @octo-patch in #1175
• Roll back task state when scheduling fails by @HaningZS in #1192
• Persist reusable WebUI generation settings by @HaningZS in #1191
• Autoplay WebUI voice previews by @AZEROSTART in #1186
• Synchronize Gemini TTS voices with the official catalog by @jxnding in #1195
• Add a Japanese README by @eltociear in #1202
• Register the ping health-check router by @NafeesMadni in #1212
• Add an Italian WebUI translation by @SandroHub013 in #1213
• Export and import WebUI settings presets and API key backups by @rignaneseleo in #1211
• Improve response schemas and migrate to Pydantic v2 ConfigDict by @NafeesMadni in #1217
• Add WaveSpeed AI as a text-to-video material source by @chengzeyi in #1201
• Add Anthropic Claude as an LLM provider by @Gainto in #1223
• Fail fast with an actionable message when FFmpeg is missing by @alvinhui in #1222
• Use LLM-generated social captions for every cross-post platform by @Kizuno18 in #1230
• Restrict custom audio file paths by @lihuiyang1024 in #1238
• Prevent task symlink traversal by @lihuiyang1024 in #1239
• Harden local material uploads by @lihuiyang1024 in #1241
• Sanitize client request IDs by @lihuiyang1024 in #1242
• Add Fish Audio TTS integration by @hariom-hp in #1228

New Contributors

• @HaningZS made their first contribution in #1192
• @AZEROSTART made their first contribution in #1186
• @jxnding made their first contribution in #1195
• @eltociear made their first contribution in #1202
• @NafeesMadni made their first contribution in #1212
• @SandroHub013 made their first contribution in #1213
• @rignaneseleo made their first contribution in #1211
• @chengzeyi made their first contribution in #1201
• @Gainto made their first contribution in #1223
• @alvinhui made their first contribution in #1222
• @lihuiyang1024 made their first contribution in #1238
• @hariom-hp made their first contribution in #1228

Full Changelog: v1.3.4...v1.3.5

---

重点更新

• 新增 MiniMax 和 Fish Audio 文本转语音服务，完善 MiniMax 区域端点和音色选择支持，同步 Gemini 官方音色目录，并在 WebUI 中增加自动语音试听功能。
• 新增 Anthropic Claude 原生大语言模型服务商。
• 新增胜算云（Shengsuan AI）和 WaveSpeed AI 文生视频素材源。
• WebUI 新增可复用的视频生成设置以及预设导入、导出功能，并支持选择性备份和恢复 API Key。
• 所有受支持的跨平台发布渠道现在都可以使用大语言模型生成社交媒体文案，不再仅限于 YouTube。
• 新增意大利语 WebUI 翻译和日语 README 文档。

安全性

• 为 /api/v1 接口和生成的任务文件增加可选的 API Key 鉴权；当 app.api_key 保持为空时，现有安装和客户端仍可继续使用，无需修改。
• 限制自定义音频文件只能从允许的目录读取，防止路径穿越。
• 防止通过符号链接访问任务目录之外的文件。
• 加强本地素材上传的文件类型、路径和文件名校验。
• 对客户端传入的请求 ID 进行清理和规范化处理，避免其影响任务文件路径和处理流程。

修复与改进

• 修复部分文字和字体组合下字幕最后一行被裁切的问题，覆盖多行字幕以及不包含下伸字符的文本。
• 当系统缺少 FFmpeg 时立即停止启动，并提供清晰、可执行的错误提示。
• 修复 Docker 依赖安装重试逻辑，确保软件包持续安装失败时终止构建，而不是静默忽略错误。
• 后台任务调度失败时自动回滚任务状态，避免任务停留在错误状态。
• 改进 API 响应模型，并将相关 Pydantic 配置迁移到 Pydantic v2 ConfigDict。
• 正确注册 Ping 健康检查接口。
• 保留大语言模型响应中的段落和换行结构。
• 在配置和 WebUI 中明确区分不同的 Kimi API 平台。
• 支持读取包含多个 UTF-8 BOM 标记的配置文件。
• 改进 MiniMax 音色发现与选择，并在替换已有音频文件前验证新生成的音频是否有效。

升级说明

• 默认情况下不会启用 API 鉴权，现有用户升级后不受影响。如需保护 V1 API 和生成的任务文件，请在 config.toml 中设置 app.api_key，并在请求中通过 x-api-key 请求头传递相同的值。
• 开启 API Key 备份后，导出的 WebUI 设置预设可能包含服务商凭据。请妥善保存导出文件，不要将其提交到公开代码仓库。
• Fish Audio、MiniMax、Anthropic Claude、WaveSpeed 和胜算云功能需要配置对应服务商的凭据后才能使用。

主要变更

• #1175：新增 MiniMax 文本转语音支持，感谢 @octo-patch。
• #1192：任务调度失败时回滚任务状态，感谢 @HaningZS。
• #1191：保存可复用的 WebUI 视频生成设置，感谢 @HaningZS。
• #1186：新增 WebUI 语音自动试听功能，感谢 @AZEROSTART。
• #1195：同步 Gemini 官方 TTS 音色目录，感谢 @jxnding。
• #1202：新增日语 README，感谢 @eltociear。
• #1212：注册 Ping 健康检查接口，感谢 @NafeesMadni。
• #1213：新增意大利语 WebUI 翻译，感谢 @SandroHub013。
• #1211：新增 WebUI 设置预设和 API Key 备份的导入导出功能，感谢 @rignaneseleo。
• #1217：改进响应模型并迁移至 Pydantic v2 ConfigDict，感谢 @NafeesMadni。
• #1201：新增 WaveSpeed AI 文生视频素材源，感谢 @chengzeyi。
• #1223：新增 Anthropic Claude 大语言模型服务商，感谢 @Gainto。
• #1222：缺少 FFmpeg 时快速失败并显示明确提示，感谢 @alvinhui。
• #1230：为所有跨平台发布渠道生成大语言模型社交文案，感谢 @Kizuno18。
• #1238：限制自定义音频文件路径，感谢 @lihuiyang1024。
• #1239：防止通过任务文件符号链接进行路径穿越，感谢 @lihuiyang1024。
• #1241：加强本地素材上传安全校验，感谢 @lihuiyang1024。
• #1242：清理客户端传入的请求 ID，感谢 @lihuiyang1024。
• #1228：新增 Fish Audio 文本转语音服务，感谢 @hariom-hp。

新贡献者

感谢 @HaningZS、@AZEROSTART、@jxnding、@eltociear、@NafeesMadni、@SandroHub013、@rignaneseleo、@chengzeyi、@Gainto、@alvinhui、@lihuiyang1024 和 @hariom-hp 首次参与 MoneyPrinterTurbo 项目贡献。

完整变更记录：v1.3.4...v1.3.5

v1.3.4

Highlights

• Added a WebUI preference to disable automatically opening the task folder after video generation.
• Generated videos now use the video subject as the download filename, with automatic numbering for multi-video tasks.
• Added configurable Whisper initial_prompt support to improve subtitle recognition for names, terminology, and language-specific vocabulary.
• Added persistent material search caching to reduce duplicate provider requests, API usage, and rate-limit pressure.
• Stock material selection now prioritizes assets matching the target video orientation.
• Added sanitized material source provenance to task artifacts for easier source tracking.

Fixes and Improvements

• Fixed WebUI configuration changes potentially blocking the page while a video generation task was running.
• Fixed the ElevenLabs API Key being cleared after a WebUI restart or browser reconnection.
• Unified ElevenLabs API Key resolution across TTS and music generation, including ELEVENLABS_API_KEY environment variable support.
• Fixed configuration updates when config.toml is provided through a bind mount.
• Improved diagnostics for Pixabay 429 responses and Cloudflare challenges.
• Prevented Pixabay API Keys from appearing in application logs.
• Added response and resource limits to ElevenLabs music requests to prevent excessive memory usage.
• API requests now reject video counts and clip durations below 1.
• Fixed Worker crashes caused by stale or invalid queued task parameters from earlier versions.
• Expanded the WebUI video subject input for easier editing of longer subjects.
• Updated project documentation and related resources.

What's Changed

• Enforce ElevenLabs resource limits by @arichazan in #1131
• Redact Pixabay API Keys from logs by @its-How in #1130
• Validate positive video generation limits by @ShiroKSH in #1149
• Handle stale queued task parameters without crashing the Worker by @kapil971390 in #1150
• Make Whisper initial_prompt configurable by @Blackspyda17 in #1163

New Contributors

• @arichazan made their first contribution in #1131
• @its-How made their first contribution in #1130
• @ShiroKSH made their first contribution in #1149
• @Blackspyda17 made their first contribution in #1163

Full Changelog: v1.3.3...v1.3.4

v1.3.3

Highlights

• Added full voiceover preview with estimated duration before video generation. Preview audio can be reused when the final generation settings remain unchanged.
• Added optional video-matched background music generation through Sonilo and ElevenLabs, with graceful fallback when music generation fails.
• Added custom background music uploads directly from the WebUI.
• Added a global video clip speed control and new ZoomIn / ZoomOut transitions.
• Added lightweight, non-blocking update notifications in the WebUI.
• Hardened task execution and recovery across the WebUI, API, and CLI, with clearer stage and error tracking.

Fixes and Improvements

• Fixed Windows portable-package startup failures caused by conflicting Python app packages.
• Fixed the WebUI becoming blank after starting video generation.
• Preserved generation progress, logs, and terminal task states across Streamlit reruns.
• Improved interrupted-task recovery and kept social publishing failures independent from successful video generation.
• Added queue limits and more reliable state updates for cross-platform publishing.
• Prevented unexpected Whisper model fallback when another subtitle provider fails.
• Fixed LLM provider errors being passed into material keyword generation.
• Fixed repeated cleanup attempts for the same temporary video file.
• Allowed video materials that fall only a few pixels below the nominal minimum resolution because of encoder rounding.
• Standardized TTS API key links and provider setup guidance.
• Completed missing secondary translations across supported WebUI languages.
• Added line-ending normalization for more consistent contributions across Windows, macOS, and Linux.
• Expanded CI coverage for task execution, API controllers, media handling, queues, application startup, and failure recovery.

What's Changed

• Expand the full CI test suite by @rongdu88668-ux in #1048
• Fix LLM term generation error handling by @techdou in #1097
• Add custom background music uploads by @pcbntt in #1101
• Add line-ending normalization by @Kizuno18 in #1103
• Add ZoomIn / ZoomOut video transitions by @Kizuno18 in #1104
• Add a global video clip speed control by @pcbntt in #1106
• Improve tolerance for slightly undersized video materials by @pcbntt in #1108
• Add optional Sonilo video-matched background music by @cindyxu1030 in #1114

New Contributors

• @rongdu88668-ux made their first contribution in #1048
• @techdou made their first contribution in #1097
• @Kizuno18 made their first contribution in #1103
• @cindyxu1030 made their first contribution in #1114

Full Changelog: v1.3.2...v1.3.3

v1.3.2

Highlights

• Redesigned the WebUI with a cleaner responsive layout, a dedicated settings dialog, improved dark mode support, and a more consistent interaction experience.
• Added task management with status filtering, progress tracking, video playback, folder access, deletion, and regeneration from historical task settings.
• Added a lightweight onboarding tour to guide new users through configuration, video options, and generation.
• Added built-in LLM connection testing so provider credentials and model settings can be verified before generation.
• Centralized LLM provider metadata, defaults, configuration tips, and registration links for more consistent setup and maintenance.
• Added video cache statistics and cleanup controls.
• Expanded the command-line workflow and added the MoneyPrinterTurbo Agent skill for automated video generation.
• Updated the project version to v1.3.2.

Fixes and Improvements

• Migrated Gemini text and TTS integrations from the deprecated google-generativeai SDK to the new google-genai SDK.
• Fixed WebUI selectors that sometimes required two selections before a change took effect.
• Preserved generation logs across Streamlit reruns and fixed logger cleanup errors after successful generation.
• Improved task state normalization so filtering works consistently across all UI languages.
• Prevented running tasks from being deleted and tightened task file operation boundaries.
• Improved audio settings with clearer narration modes, provider-specific setup guidance, and more consistent preview controls.
• Fixed Azure TTS V2 preview generation so the selected speech rate is applied correctly.
• Improved subtitle settings, background controls, contrast warnings, default restoration, and multilingual rendering reliability.
• Disabled subtitle backgrounds by default while retaining optional rounded translucent backgrounds.
• Added automatic video encoder selection with safer fallback to libx264.
• Improved video duration handling with a small safety margin to prevent incomplete output.
• Fixed duration detection for custom audio formats other than MP3.
• Improved LLM provider defaults, descriptions, registration links, and multilingual configuration guidance.
• Removed deprecated or unsupported provider and video source options.
• Improved WebUI performance by caching font and background music discovery.
• Updated Streamlit and related dependencies, Docker configurations, startup scripts, and toolbar behavior.
• Refined multilingual UI copy and layout across all supported languages.
• Updated README documentation, screenshots, gallery previews, configuration examples, and the Colab notebook.
• Expanded regression coverage for WebUI state, task management, providers, CLI workflows, subtitles, caching, and configuration.

What's Changed

• Fix custom audio duration handling for non-MP3 files by @pcbntt in #1099
• Add a video duration safety buffer by @Sushanth012 in #1003

Full Changelog: v1.3.1...v1.3.2

v1.3.1

Highlights

• Added new LLM providers and integrations, including AIML API, EvoLink, and VolcEngine Ark.
• Added new TTS options, including ElevenLabs and self-hosted OpenAI-compatible Chatterbox TTS.
• Added optional TwelveLabs video AI integration for smarter material understanding and reranking.
• Added YouTube Shorts publishing support through Upload-Post.
• Added a pure CLI workflow and exposed more WebUI-backed video parameters to command-line generation.
• Added GHCR Docker image publishing for easier container deployment.
• Improved multilingual WebUI support, including Indonesian and Spanish updates, plus explicit es-ES script generation support.

Fixes and Improvements

• Fixed subtitle background clipping and improved subtitle rendering reliability.
• Fixed custom-audio subtitle generation when using Whisper.
• Fixed Windows MoviePy temporary-audio handling to avoid Windows Defender interference.
• Fixed Streamlit first-run prompt blocking WebUI startup.
• Fixed LLM provider selectbox switching that sometimes required selecting twice.
• Fixed video duration safety margin handling.
• Fixed video_contact_mode typo compatibility by correctly using video_concat_mode.
• Fixed Pollinations default API endpoint.
• Improved robust JSON parsing for fenced or multi-line LLM responses.
• Exposed match_materials_to_script in the terms API.
• Improved material matching to better follow script order.
• Hardened CLI argument validation for local video material workflows.
• Redacted LLM error output to avoid leaking sensitive details.
• Updated sponsor acknowledgements and provider copy.

What's Changed

• Add pure CLI workflow for headless video generation by @copilot-swe-agent in #1005
• Fix video_contact_mode typo and use video_concat_mode by @reyys in #1021
• Add AIML API provider by @D1m7asis in #1022
• Fix Pollinations default endpoint by @muhamadmaliqia in #1025
• Fix robust JSON parsing from fenced / multi-line LLM responses by @xlab-nyc in #1026
• Fix MoviePy temporary-audio handling on Windows by @mertfurkandgn in #1029
• Prevent Streamlit first-run prompt from blocking WebUI startup by @BinaryOutlook in #1030
• Add CI and targeted quality fixes by @BinaryOutlook in #1031
• Validate --video-source local constraints in CLI by @kapil971390 in #1033
• Port GUI-backed customization options to CLI by @BinaryOutlook in #1035
• Add Spanish (Spain) WebUI translation by @lsopenad in #1037
• Add YouTube Shorts support through Upload-Post API by @lsopenad in #1046
• Expose match_materials_to_script in the /terms endpoint by @kapil971390 in #1065
• Add Indonesian WebUI translation by @mteguhfh in #1066
• Add ElevenLabs TTS support by @nhatanh2996 in #1071
• Add Chatterbox TTS provider by @pcbntt in #1077
• Fix YouTube Shorts metadata upload by @kapil971390 in #1078
• Add optional TwelveLabs video AI integration by @mohit-twelvelabs in #1084

New Contributors

• @copilot-swe-agent made their first contribution in #1005
• @reyys made their first contribution in #1021
• @D1m7asis made their first contribution in #1022
• @muhamadmaliqia made their first contribution in #1025
• @xlab-nyc made their first contribution in #1026
• @mertfurkandgn made their first contribution in #1029
• @BinaryOutlook made their first contribution in #1030
• @lsopenad made their first contribution in #1037
• @mteguhfh made their first contribution in #1066
• @nhatanh2996 made their first contribution in #1071
• @pcbntt made their first contribution in #1077
• @mohit-twelvelabs made their first contribution in #1084

Full Changelog: v1.3.0...v1.3.1

v1.3.0

Highlights

• Added Coverr as a new stock video material provider, including WebUI API key management and 16:9-friendly defaults.
• Added Groq as a new LLM provider.
• Added no-voice video generation mode for workflows that do not need narration.
• Added configurable video encoder selection and safer hardware-encoder fallback behavior.
• Added AIHubMix provider integration.
• Added multilingual social metadata generation API.
• Added subtitle background options, including optional rounded translucent subtitle backgrounds.
• Improved WebUI localization coverage, including Arabic, German, Russian, Turkish, and Vietnamese updates.

Fixes and Improvements

• Improved repeated video material reuse behavior.
• Fixed final SRT subtitle block handling.
• Fixed subtitle matching for markdown-style script markers.
• Fixed Qwen response parsing edge cases.
• Upgraded and hardened selected dependencies and deployment defaults.
• Refactored Azure voice data into JSON for easier maintenance.
• Improved README and setup documentation quality.

What's Changed

• refactor(voice): extract Azure voice data to JSON by @domantasjo in #951
• fix: use Optional[list] type annotation in upload_post.py by @kuishou68 in #960
• i18n(vi): complete Vietnamese UI translation (135/135 keys) by @luongs3 in #963
• docs(README-en): remove obsolete ImageMagick requirement and fix doc quality issues by @jonathanzhan1975 in #977
• fix: keep last subtitle when SRT has no trailing blank line by @Bortlesboat in #987
• fix: handle empty Qwen choices by @Sushanth012 in #994
• i18n: complete Russian WebUI translation by @DrMaks22 in #996
• feat: add Groq support by @RafaGomezGuillen in #997
• fix(groq): add default model name + clarify unvalidated manual entry warning by @kapil971390 in #1014
• feat: add Coverr video material source by @shoujche in #1008

New Contributors

• @domantasjo made their first contribution in #951
• @kuishou68 made their first contribution in #960
• @luongs3 made their first contribution in #963
• @jonathanzhan1975 made their first contribution in #977
• @Bortlesboat made their first contribution in #987
• @Sushanth012 made their first contribution in #994
• @DrMaks22 made their first contribution in #996
• @RafaGomezGuillen made their first contribution in #997
• @kapil971390 made their first contribution in #1014
• @shoujche made their first contribution in #1008

Validation

• uv lock --check
• uv run --with pytest pytest test/services -q
• Result: 117 passed, 1 skipped, 18 warnings, 9 subtests passed.

Full Changelog: v1.2.9...v1.3.0