AutoCut Updates

AutoCaptions gets a major upgrade to make caption editing faster, safer, and more precise, especially for social-first workflows.

Chunk preview (what you see is what you get)

Each caption chunk now includes a live preview of how it will appear in the video. No more guessing how captions will render : adjust timing, text and style with full visual confidence.

Safe area grid for social formats

A new safe area grid is now available, adapting automatically to major social formats. This helps you keep captions perfectly readable and avoid UI cutoffs on TikTok, Reels, Shorts and more.

Listen to your transcript

You can now play the transcript directly from AutoCaptions, with adjustable playback speed (x1, x2, and more). Proofread, review, and fine-tune captions while listening without leaving your editing flow.

Improved captions presets listing

The captions presets library has been redesigned for better readability, discovery, and consistency. Community presets are now easier to browse, compare and reuse across projects.

This update makes AutoCaptions more visual, more social-ready and more editor-friendly, while keeping full creative control inside your timeline.

AutoCut and the entire website are now fully available in German and Italian.

This update offers a smoother, more accessible experience for thousands of creators across Europe and brings us one step closer to making AI-powered editing available to everyone.

AutoB-Rolls now includes native Storyblocks integration, allowing you to browse, select and insert premium stock footage directly from within AutoCut. No external search, no downloads, no imports.

This update delivers a faster, more coherent B-roll workflow and significantly improves visual quality across your edits.

Learn how to use Storyblocks inside AutoCut here

AutoCaptions Improved handling of quotation marks AutoCut now detects and formats quotes correctly across all supported languages. Citations, dialogues and quoted sentences are finally transcribed with clean, reliable punctuation.

Smarter accent processing Words containing accents are now automatically normalized whenever a font cannot display them properly. This ensures consistent captions in fast-paced edits and avoids broken characters on social formats.

New “Correct with AI” button A new button inside the Captions panel allows you to fix transcription mistakes instantly. It analyses the text and corrects spelling and formatting automatically.

AutoB-Roll

New generation models We introduced updated image generation models that produce more detailed, more realistic and more coherent visuals. This upgrade improves every automatically generated B-roll, whether it's used for tutorials, talking heads or social formats.

Autofill for all slots You can now generate every B-roll in a single click. All slots are filled automatically, speeding up the process and making batch creation more consistent. This reduces back-and-forth and gives a much faster creative experience.

AutoViral

New color-coded markers AutoViral now introduces makers that make it easier to identify viral moments inside long videos. These markers provide a clearer reading of the timeline and allow you to spot high-impact segments at a glance.

AutoCut Silences — New preview A new preview experience is now available for AutoCut Silences, making pause detection clearer and more interactive.

AutoCaptions — More accurate & more expressive

• Bigger Font animation A new “Bigger Font” animation style is now available — perfect for punchy, high-impact captions.

• New transcription model We introduced a more accurate transcription model, especially for the languages shown below:

• Translation fix Fixed an issue where some caption chunks were not translated properly.

AutoProfanity — More control

• Word groups You can now add multi-word groups (e.g., “big trouble maker”) for more flexible detection.

• Residual audio fix A bug causing some censored words to still be heard has been fixed.

Other improvements Various internal updates and fixes to improve stability and accuracy across the plugin.