Product Updates

• Improved Language Detection and Manual Language Selection: Sharper detection accuracy for heavily accented speech, regional dialects, and less commonly supported European languages, resulting in more reliable transcription and caption timing across a wider range of real-world creator content. Creators can also now manually select the source language before processing - useful for content where auto-detection may be uncertain or where a specific language variant is preferred.
• More Reliable YouTube Imports: YouTube imports are now significantly more stable on slower or intermittent connections, with better retry handling and timeout management that reduces failed imports without requiring manual retries.
• Offline Resilience: Reelify no longer requires an active server connection to remain functional. During scheduled maintenance or unexpected downtime, creators can continue working without interruption, Reelify gracefully handles unavailable backend services instead of blocking the workflow.
• More Reliable Update Notifications: Update notifications are now delivered more consistently across all users, so every creator is promptly aware when a new version is available.

v5.3.6 August 9, 2026

• On-Device AI Clipping: After the initial AI engine download, the core clipping experience now runs entirely on-device with no internet connection required. Creators can import, analyze, and generate clips completely offline, with the same accuracy and speed as the standard workflow. Designed for creators working on the go, in low-connectivity environments, or anyone who prefers keeping their content fully local. Available on the Creator plan / lifetime deal.

v5.3 July 30, 2026 UTC

• Visual Hooks: A new retention-focused feature that adds a customizable opening title card to clips. Designed to grab viewer attention within the first few seconds and reduce early drop-off - a proven short-form technique used across TikTok, Reels, and Shorts to increase watch time and engagement.
• Improved Layout and Framing Mode Guidance: Clearer interface and messaging across custom crop, dual crop, and split-screen workflows - with full-frame coverage replacing black bars, more customizable layout controls, and a more intuitive setup flow that reduces configuration guesswork across all framing modes.
• Refined Multi-Scene Framing: The scene-based framing system introduced in v5.2 has been further improved with more consistent behavior across longer clips with frequent scene changes, tighter detection of scene transitions, and smoother framing adjustments between segments.

v5.2 July 18, 2026 UTC

• Word-Level Clip Editing from the Transcript: Trim, extend, or adjust clip boundaries with word-by-word precision directly from the transcript. Instead of dragging timeline handles or stepping through frames, creators can now click individual words to set exactly where a clip starts and ends - with the edit snapping precisely to spoken content. This is the first major milestone toward complete transcript-driven editing across the entire workflow.
• Scene-Based Framing Editor: A dedicated editor for setting different framing methods across different scenes within a single clip. For recordings where camera angles shift, scenes change, or new subjects enter the frame, creators can now assign separate framing behavior to each section individually - ensuring every part of the clip is framed correctly without compromising the rest.
• Improved Custom Font Import: The TTF font import workflow has been refined for a smoother, more reliable experience - with better handling of font weights, character sets, and rendering consistency across caption styles.
• Adjustable Brand Kit Logo Sizing: Brand logos can now be resized freely within the Brand Kit - larger or smaller depending on the content - instead of being locked to a single fixed size.

v5.1 July 14, 2026 UTC

• Modern Caption Styles: A new set of engaging, TikTok and Reels-inspired caption animations - including bounce, glow, and dynamic motion effects designed for high-engagement short-form content. This is the first phase of a broader caption overhaul that will continue expanding with additional styles and animation options in upcoming releases.
• Expanded Caption Customization: Larger caption sizes, new size presets, and significantly more control over how captions look on screen. Creators can now fine-tune caption appearance to match their content style without being limited to a small set of default options.
• Custom Font Support in Brand Kit: Upload and use your own TTF font files directly in captions through the Brand Kit. Creators and agencies working with brand guidelines can now maintain full typographic consistency across every clip without workarounds or third-party tools.
• Focused Auto-Framing Mode: A new dedicated framing mode optimized for podcasts and single-speaker recordings. Keeps the primary subject consistently centered and properly framed throughout the clip, even during movement, gestures, and camera shifts - without requiring manual adjustment.

v5.0 July 6, 2026 UTC

• Refreshed App Design: A comprehensive visual refresh across the entire interface - with a cleaner layout, refined typography, updated UI elements, and a more polished look throughout every screen.
• Support for Long-Form Videos Up to 6+ Hours: Full-length podcasts, webinars, lectures, and interviews - even those running four to six hours or longer - are now handled through intelligent segment processing. Longer videos are automatically split into optimized segments so AI clipping, captioning, and export remain accurate and stable regardless of total video length. Creators can import extended content without worrying about processing limits or degraded output quality.
• Batch Export in Original Format: Multiple clips can now be exported simultaneously in their original source format and resolution. Creators working with large clip sets can queue and render entire batches without manually configuring each export individually - significantly faster for high-volume workflows like podcast repurposing, course content, and agency-scale production.
• Multilingual Caption Improvements: Caption accuracy and word-level timing have been further refined with a particular focus on Asian language output - Chinese, Japanese, Korean, and Hindi captions are noticeably cleaner, with tighter synchronization on fast speech and more reliable handling of mixed-language segments. Improvements also extend across all other supported languages without requiring any manual configuration.

v4.6.4 July 1, 2026 UTC

• Dedicated AI Clipping Model for Asian and Global Languages: A newly integrated language-specific model now drives AI clip detection and transcription accuracy for Chinese, Japanese, Korean, Russian, and Ukrainian. By processing language structures natively, highlight identification, clip timing, caption quality, and overall clip relevance now perform at a level much closer to the English pipeline. Early support is also expanding to additional languages including Thai, Arabic, Hebrew, and Turkish, with further refinement shipping as testing continues. This is the next major step in a broader multilingual effort to bring high-quality AI clipping to every language Reelify supports.
• Transcript-Based Clipping Across 99+ Languages: The transcript browsing and search features introduced in v4.6.2 now work across all 99+ supported languages - including Hindi, Vietnamese, Indonesian, Malay, and every major Asian and Middle Eastern language. Creators working in any supported language can now browse the full transcript, search by keyword, and manually add clips from any point in the recording without relying on AI suggestions alone.

v4.6.2 June 21, 2026 UTC

• Significantly Improved AI Clipping and Transcription for European Languages: AI clip detection, transcription accuracy, and caption quality have all been substantially improved across major European languages - including Dutch, German, Norwegian, Swedish, French, Spanish, Italian, and Polish. Highlight identification, clip timing, word-level caption accuracy, and overall clip relevance are all noticeably stronger, bringing the non-English experience much closer to the accuracy creators already get with English content. This is the first major result of a broader multilingual effort that will continue expanding across future releases.
• Transcript-Based Manual Clip Creation: Browse the complete transcript directly inside the app and manually add clips from any point in the recording. Instead of scrubbing through the timeline or relying solely on AI-generated suggestions, creators can now read through what was said, find the exact moment they want, and turn it into a clip with a single action. Particularly useful for interviews, educational content, and long-form recordings where the creator already knows which sections are worth pulling.
• Full Transcript Search: Search through the entire transcript by keyword or phrase to instantly locate specific moments across a recording. Whether looking for a particular topic, a guest's name, or a key talking point buried deep in a long video - search makes it possible to jump directly to the right moment and create a clip from it, without reading or scrolling through the full transcript manually.
• More Stable YouTube Imports: YouTube URL imports are now more reliable and consistent across a wider range of video types, lengths, and upload configurations. Import success rates have been improved with better handling of edge cases around regional availability, age-gated content, and non-standard video formats - reducing failed imports and retries for creators pulling content directly from YouTube.

v4.5 June 11, 2026 UTC

• Significantly Richer Clip Refresh Suggestions: The Refresh Suggestions button now produces a noticeably wider variety and more clips on every pass - different moments, different lengths, and different angles into the same content. Each refresh pulls from a broader range of the timeline, so creators can cycle through genuinely different clip ideas without re-analyzing the video.
• More Clips per Analysis: The AI now surfaces significantly more clips from a single analysis. Longer videos produce up to 15–16 clips, and videos over one hour can return up to 20 - up from the previous ceiling of around 10. A much larger starting set to choose from, especially on podcasts, interviews, and webinars.
• Deeper Detection for Videos Over 1 Hour: Long-form content (1hr+) now receives the same depth of clip detection and timing precision as shorter videos. Highlight identification across interviews, webinars, podcasts, and extended recordings is noticeably more thorough, with better moment selection throughout the full timeline.
• Easier-to-Discover Editing Controls: Many of Reelify's most powerful features - auto-framing configuration, dual split-crop setup, clip timing adjustments, manual clip creation, and caption positioning - were already built in, but not always surfaced prominently enough. This update restructures how they are accessed, making the full editing toolkit easier to navigate and faster to reach.
• Updated Tooltip Messages and In-App Guidance: Sharper, more specific tooltip messages and contextual hints across the editing workflow - covering framing modes, crop configuration, export options, and caption setup. Each message is written to explain what a setting does and when to use it, reducing the learning curve for both new and returning users.
• Subtitle Export for Easy Repurposing: Exported subtitle files can now be used directly for writing video titles, descriptions, blog posts, or social captions - making it faster to repurpose content across platforms without manually transcribing or copying text from the editor.

v4.4.2 June 9, 2026 UTC

• Enhanced Auto-Framing Engine: Extended the auto-framing engine's accuracy and consistency across a wider range of content types — including single-subject recordings, older footage, compressed uploads, and low-resolution source files. Subject detection is tighter, crop stability is improved, and the underlying engine is being actively refined with further improvements shipping in upcoming releases.
• Extended macOS 15 Stability Optimizations: Expanded stability optimizations across older macOS 15 builds, with improved session reliability throughout the import and export workflow on pre-15.2 configurations.
• Broader Export Format Coverage: Broadened export format coverage with faster processing and tighter consistency across MKV, MOV, and variable frame rate sources — ensuring reliable output across a wider range of video containers and configurations.

v4.4 May 30, 2026 UTC

• Manual Clip Selection via Timestamps: Creators can now define custom clips by entering precise start and end timestamps directly, adding a full manual control layer alongside AI-generated suggestions. Perfect for creators who already know exactly which moments to pull — interviews, highlight compilations, or any workflow where editorial instinct should drive the cut list from the start.
• Smarter AI Clip Detection for Short-Form Source Videos: Completely reworked the AI highlight detection pipeline for source videos under seven minutes. The model now evaluates short-form content with a purpose-built scoring window tuned specifically for condensed footage, producing tighter, more relevant clip suggestions that consistently surface the highest-signal moments.
• Refined Auto-Framing for Single-Subject Content: Elevated the auto-framing engine's precision when a single subject is present in the frame. Tracking now locks onto the primary subject with tighter crop boundaries, reduced drift, and smoother repositioning during natural head movement — delivering noticeably more stable, broadcast-quality framing across solo-speaker recordings, tutorials, and talking-head content.
• Persistent Clip Edits: Clip-level modifications — including trim and extend adjustments, manual crop positioning, and Dual Stack layout configurations — are now fully retained as you move between clips and screens during your editing workflow. Every editorial decision stays locked in exactly as configured, keeping the path from first cut to final export seamless and uninterrupted.
• Caption Export as SRT & TXT per Clip: Each individual clip's captions can now be exported as standalone .srt or .txt files alongside the video output. Useful for uploading captions separately to platforms, archiving transcripts, or handing off subtitle files to editors and translators working downstream.
• Filler Word Removal from Captions (Alpha): Introducing an early-access version of automatic filler word detection and removal within the caption layer. Words like "um," "uh," "like," and other verbal fillers are identified and can be stripped from the visible caption output — producing cleaner, more polished on-screen text without altering the underlying audio track. This is the first iteration of a broader caption intelligence system, with continued refinement shipping in upcoming releases.
• Auto-Pause on Edit Navigation: Playback now pauses automatically when navigating from the clip viewer to the editor screen, keeping the editing environment focused and distraction-free while configuring crop, caption, or layout settings. A streamlined quality-of-life improvement informed directly by creator workflow feedback.
• Watermark Export Pipeline Hardened: Strengthened the watermark rendering and license validation layer across all export paths, ensuring consistently clean, watermark-free output for all active paid plan holders regardless of export format or aspect ratio configuration.

v4.3.1 May 22, 2026 UTC

• Original Aspect Ratio Exports: Clips can now be exported at their native source resolution and aspect ratio — including 16:9 landscape and 1:1 square — in addition to the standard 9:16 vertical output. This expands Reelify's export coverage to YouTube repurposing, tutorial distribution, social feed posts, and any workflow that goes beyond vertical-only delivery.
• Caption Engine Accuracy & Workflow Improvements: Delivered a focused accuracy pass on the transcription layer, reducing word-error rate on fast speech and domain-specific vocabulary. Alongside that, completely reworked the caption editing experience — clearer contextual messaging, smarter inline hints, and a more structured editing flow that shortens the time from raw transcript to polished, export-ready captions.
• Clips Folder Access Directly from the App: Added a one-click shortcut to open the saved clips output folder directly from within the app. Removes the need to navigate through Finder after every export and keeps the post-export handoff workflow noticeably faster.
• Per-Clip Output Location: Users can now choose where each individual clip is saved at export time, rather than relying on a fixed default path. Particularly useful for creators managing multiple projects or delivering files directly to client-specific folders.
• Refined Auto-Framing & Split-Framing Mode Guidance: Redesigned the in-app contextual guidance for both Auto Framing and Split Framing modes — richer explanations, clearer setup steps, and sharper differentiation between the two workflows. Reduces configuration errors and lowers the learning curve for both new and returning users switching between modes.
• Improved Auto-Framing Accuracy: Tightened face detection responsiveness and tracking consistency across a broader range of video types, reducing common drift patterns and frame instability during subject movement and scene transitions.

v4.2.1 May 19, 2026

• Caption Vertical Positioning: Added vertical placement controls to the caption editor, allowing creators to set exactly where captions sit in the frame. Move them higher or lower to avoid covering key on-screen subjects or platform UI elements on TikTok, Reels, and Shorts.
• Persistent Style Presets & Auto-Caching: Saved styling preferences are now retained across editing sessions. Customize your fonts, weights, shadow backdrops, and active highlight colors once, and they will load automatically for every subsequent project without requiring manual reconfiguration.
• Enhanced Multi-Speaker Layout Flow: Redesigned the user experience and configuration panels for both Dual Stack Crop and Split Crop modes, making it significantly easier to define separate cropping parameters for multi-host and reaction layout formats.
• More Accurate Caption Preview: Caption preview in the editor now more closely matches the final exported output, reducing the gap between what you see while editing and what renders to the file.
• Improved Export Reliability: Strengthened the export pipeline to handle back-to-back rendering sessions more consistently across all device configurations.

v4.1.2 May 14, 2026

• Improved guidance and system messaging for non-English users, with clearer support for multilingual caption setup, language-specific workflows, and caption-related status updates.

• Rebuilt caption engine with significantly higher word-level accuracy. Handles fast speech, overlapping audio, and domain-specific vocabulary more reliably than before.
• Caption preview in the editor before export. See exactly how font, size, position, and style will look on a clip before rendering begins.
• New modern white caption styles with clean, high-contrast backgrounds. Well suited for professional and corporate content.
• Estimated export time shown per clip before the queue starts, so batches are easier to plan around.
• Clearer messaging and guardrails when working with large export files or uncommon video containers. Actionable guidance surfaces instead of vague or silent failures.
• New dual stack crop mode for PiP, reaction, and interview-style videos. Stack two speaker feeds vertically with independent crop controls per track, no manual timeline splitting required.
• Substantially improved auto-framing for long-form clips. Face tracking now holds stable across scene cuts and multi-speaker segments with far less drift.
• Resolved reliability issues causing exports to stall, produce corrupted output, or drop audio sync on clips longer than 10 minutes.

v3.8.1 April 27, 2026

• Redesigned caption editor with finer per-word timing and position controls.
• Expanded the built-in font library with new premium typefaces across all caption styles.
• Significantly improved framing accuracy in AI split-screen mode, particularly for side-by-side and reaction layouts.
• Minor UI refinements and stability fixes across the export pipeline.

v3.7.8 April 15, 2026

• AI clipping now supports 99+ major languages. Highlight detection works across non-English content with the same accuracy as English.
• Auto-captions extended to 99+ languages. Transcription and word-level sync work end-to-end without manual configuration.
• Brand Kits added. Set your fonts, colors, and logo once, and Reelify applies them consistently across every clip you export.

v3.7.6 April 6, 2026

• Expanded caption customization with new controls for line height, max characters per line, and per-word color overrides.
• Improved automated caption positioning. Captions now avoid covering key visual elements more reliably across different aspect ratios.
• New caption visual styles and color palettes added to the style picker, including several high-contrast and gradient options.

v3.7.2 April 1, 2026

• Streamlined editing interface with a reorganized toolbar and reduced click depth for common actions.
• Improved compatibility with the iOS companion app, including faster project sync and fewer handoff errors.

v3.6.9 March 26, 2026

• URL imports are significantly more reliable. Addressed the most common failure modes caused by platform-side changes on YouTube and Vimeo.

v3.6.7 March 22, 2026

• Added a new set of high-impact caption styles optimized for short-form content and high viewer retention.
• Blur mode exports are noticeably faster on Apple Silicon, with reduced thermal load on longer renders.

v3.6.5 March 16, 2026

• Manual crop and split-screen layout launched in beta. Set custom crop regions for each speaker track independently before exporting.