MoneyPrinterTurbo

English | 中文

• 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

English | 中文

• 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 supp

English | 中文

• 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](https://github.com/harry0703/MoneyPrinterTurbo/commit/9f0b28f8e87db76feee2d49ad3d98a31b43

English | 中文

• 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 https://github.com/harry0703/MoneyPrinterTurbo/pull/1246
• Add container provenance and SBOM attestations by @kobihikri in https://github.com/harry0703/MoneyPrinterTurbo/pull/1140
• Add the Upload-Post Auto-Publish WebUI settings by @parveen0029 in https://github.com/harry0703/MoneyPrinterTurbo/pull/1209
• Improve headless WebUI video preview behavior by @abhiunix in https://github.com/harry0703/MoneyPrinterTurbo/pull/1252
• Apply saved WebUI settings as CLI defaults by @Gainto in https://github.com/harry0703/MoneyPrinterTurbo/pull/1225
• Decouple Upload-Post configuration from automatic publishing by @mcgeand in https://github.com/harry0703/MoneyPrinterTurbo/pull/1259
• Preserve logs across different filesystem mounts by @SandroHub013 in https://github.com/harry0703/MoneyPrinterTurbo/pull/1260
• Add CLI batch manifest mode by @lihuiyang1024 in https://github.com/harry07