What we're building next

Swipeless tops out at roughly 30 minutes per video. The word target in the assistant caps at 5000 words, and "match the source video's length" caps at 4000, which works out to about 32 and 26 minutes of narration. That rules out the compilation formats some niches are built on, where a single upload runs two to four hours. Raising the ceiling is not one number: the word caps, the batch ceiling for image generation and the assembly timeout all move together, and a three hour video means around 1900 scenes and roughly 55 to 60 dollars of cloud spend per upload. A cheaper first step may be joining several finished projects into one file, which does not exist today either.

The Projects page always opens on "All", so anyone who keeps several channels apart with categories re-picks the same one every visit. This remembers the filter you last used, including "Uncategorized", and opens there next time.

The "animate the first N seconds" choice is made before the scenes are cut, so a project that ran on a normal template can never be moved into it afterwards. The seconds box in the Animate step only appears when the project already carries an opening window. Let it be chosen later too, including right after an automation run has finished and a few images have been corrected.

The Gemini voices have no preview in the app. Only the stock ElevenLabs voices do, so the only way to hear one is to generate a whole voiceover. Add a play button on every voice: one short sample per voice, generated once and reused for everyone, free to listen to.

Connect Swipeless to your AI assistant via the Model Context Protocol. Tigger video generation directly from your terminal or IDE using natural language commands.

An autopilot uploads a finished video the moment its production run ends, so the publish time drifts with however long the render took. This adds an optional publish time to each autopilot: pick an hour and minute in your timezone, and finished videos wait in the queue until that slot instead of going out immediately. A video that finishes after its slot goes out at the next one, so a slow render shifts a day rather than skipping it. Requested by a customer who wants a consistent daily upload time on their channel.

Invite your team to join your account so you can manage, edit, and produce videos together on the same projects.

This feature adds support for Claude and OpenAI models to assist in automated script generation.

This feature enables users to connect their social media accounts and automatically publish videos to Facebook, Instagram, X, and YouTube. It helps content creators save time by streamlining their distribution workflow across multiple platforms.

This feature enables users to create content in a 9:16 vertical aspect ratio. It helps creators produce mobile-optimized videos that align with modern social media formats.

This feature adds Kling and Seedance as selectable options for video and image generation within the platform. It provides users with more creative flexibility and advanced model choices to enhance their visual content production workflows.