This week in AI video and filmmaking – September 6, 2026

New models, tools and platform rules from the past week in AI video and AI filmmaking. Every entry names its source and the date we checked it.

• New model: PixVerse Adds Seedance 2.5 With Video and Transition Modes
  PixVerse has added Seedance 2.5, supporting Video, Transition, and Reference modes with clip lengths from 4 to 30 seconds. This gives PixVerse users a newer Seedance version alongside the earlier 2…
• Update: Runway Update Adds Native Audio, Lip-Sync and API Hooks
  A May 2026 Runway update introduced native audio generation, lip-sync, social media templates, and new API hooks for developers. The changes expand what creators can produce and automate directly o…
• Pricing: Synthesia Raises Creator Plan Price, Adds Recurring Free Tier
  Synthesia increased its Creator plan to $89 per month while introducing a recurring free tier offering 10 video minutes monthly. The price change affects paid users while the free tier gives new us…
• New tool: Higgsfield Launches Genjutsu Video Editor for Object and Scene Swaps
  Genjutsu is now available on Higgsfield, letting users change objects, locations and characters in existing videos using Motion Transfer and Object Swap modes. The tool is live as of September 1.
• New model: Higgsfield Adds Recraft V4.1, Styles and Utility Models
  Three Recraft V4 models are now available on Higgsfield: V4.1 for lighting, Styles for style matching, and Utility for simpler outputs. All three launched on August 27.
• New tool: Higgsfield Adds FLUX 3 Video Upscaler for Up to 4K Output
  The FLUX 3 Video Upscaler is now live on Higgsfield, letting users raise video resolution to 1080p, 2K or 4K. It launched on August 27.
• New model: InVideo AI Integrates Google Veo 3.1 for Text-to-Video Generation
  InVideo AI has added Google’s Veo 3.1 model, giving users native text-to-video generation inside the platform.
• Platform rules: YouTube Shorts Adds Speech to Song Using Lyria 2 Music Model
  YouTube Shorts now supports AI-generated video creation with a Speech to Song feature powered by Google’s Lyria 2 music model, letting creators turn spoken audio into song within Shorts.

Read the full feed on GraiLogic