Roadmap | Rewind.ai

This page divides work into three groups: Completed covers shipped features, In Progress covers what the team is actively building right now, and Planned covers what comes next. Check back after any major release to see items move between groups.

Three features are under active development: voice cloning (reproduce any voice from a short sample audio clip), 3D model generation from text or image input, and PDF chat (upload a document, then ask questions about it).

The planned list includes an AI podcast generator, talking head video avatars, real-time voice chat, dark mode across all pages, and a Progressive Web App so you can install Rewind.ai on a phone or tablet from the browser.

Yes. Use the contact link at the bottom of this page to submit a request. Priority is based on how many users have asked for something, how feasible it is to build with open-source tooling, and how well it fits the platform.

The page updates whenever a feature changes stage. A finished in-progress item moves to completed and a note about its implementation is added to the changelog. New planned items are added as user feedback and development capacity allow.

Voice cloning will use the same token model as the other tools. Free users will get a daily quota, and paid plans will raise that limit. The underlying models carry permissive licenses (Apache 2.0 or MIT) so there are no licensing restrictions on personal use.

PDF chat lets you upload a document and then ask questions about it in plain text. The AI extracts and indexes the content so you can get a summary, locate a specific detail, or ask for an explanation of a section without reading the whole file.

Dark mode is listed under Planned. It will cover every tool interface and page rather than just some of them. There is no release date yet, but it is one of the most-requested items and is treated as a priority.

The planned podcast generator takes a topic or outline and produces a full episode: scripted dialogue, multiple synthesized voices and mixed audio that can be exported and published directly. It will rely on the same text-to-speech and writing models already live on the platform.

A Progressive Web App (PWA) is on the planned list. You install it from your mobile browser and it behaves like a native app: it appears on your home screen, can work offline for certain tasks, and can receive push notifications.

The team reviews new open-source model releases on an ongoing basis. The main filters are license terms (Apache 2.0 or MIT preferred), quality on published benchmarks, and whether the output is actually useful for real tasks rather than just scoring well.

The roadmap is forward-looking: it shows what is planned and what is being built. The changelog is backward-looking: it records what has shipped and when. When something moves from In Progress to Completed here, the changelog gets a corresponding entry with implementation details.