Feeds in. Finished videos out.
Point PostWire at a feed and it writes the script, records the voiceover, renders the video and publishes it to your channels. You approve every one first.
PostWire is a video automation platform for creators and marketing teams. It reads content from a source you choose, generates a narrated video from it with AI, and publishes that video to your own TikTok, YouTube, Instagram and Facebook accounts on a schedule you set.
The work it replaces is production work, not creative work. Finding the story, recording the voice, cutting the captions, exporting the file and remembering to post it. PostWire runs that part as a pipeline, so the only thing left on your desk is the judgement call.
A feed, a subreddit, a YouTube playlist or your own notes. Items deduplicate by content, so a story never becomes two videos.
A script in your words, spoken aloud, captioned line by line and rendered at every size you asked for.
The run stops and waits with the finished video. Nothing reaches an audience before you say so.
One job per channel. If one platform fails the others still ship, and the run says so plainly.
A source can run its items past a model before anything is made from them. Reject what does not belong, score what is worth making, rewrite a feed full of ticker symbols and abbreviations so it reads properly when spoken aloud. You write the instruction; it runs on every fetch.
A 9:16 cut for TikTok and a 16:9 cut for YouTube are the same video rendered twice. Same story, same script, same voice. PostWire renders them together, and each cut publishes to its own channels.
Long-form works the same way, including the case that pays for itself: point a pipeline at the shorts you already made and it compiles them into one long video, so a month of short-form spend produces something else at the end of it.
Each step records what it did, how long it took and what it cost. When something fails you get the step, the reason and the retry, not a red banner.
Tokens are held when a run starts and settled when it ends. The ledger shows the charge per step and the model that earned it.
New pipelines are review-first. Auto-publish exists, but you turn it on deliberately, per pipeline.
Every instruction the AI receives is yours to rewrite, and every step picks its own model. A daily explainer and a weekly deep dive do not have to cost the same, or sound the same.
Install PostWire on a Mac or PC and it renders on your hardware instead of ours, for a fraction of the tokens. If you already run ComfyUI or Ollama it drives those too - it ships no models and no Python, so you keep the setup you have.
The expensive step runs on your machine. Scheduling, storage and delivery stay with us, and that is all you are charged for.
Every run reserves what it would have cost us to render. Using your own machine can only ever settle lower than the estimate you approved.
The app carries the render agent inside it. Sign in and the machine starts taking work - no pairing code, no terminal, no second thing to set up.
Checking for a build…
One balance of platform tokens covers every pipeline. A run holds an estimate up front, then settles against real usage when it finishes, so an abandoned draft does not quietly bill you.
2,000 tokens included
Enough to build a pipeline and watch it produce real videos before you decide anything.
PostWire is a video automation platform for creators and marketing teams. It reads content from a source you choose, generates a narrated video from it with AI, and publishes that video to your own TikTok, YouTube, Instagram and Facebook accounts on a schedule you set.
Only if you tell it to. Pipelines default to review-first: the run stops at an approval gate with the finished video and waits. Auto-publish is a per-pipeline setting, not the default.
RSS and Atom feeds, Reddit, YouTube playlists, videos you have already made with PostWire, and stories you paste in yourself. Sources deduplicate by content, so the same story never becomes two videos.
Yes. A pipeline can render any frame size, and one run can produce several cuts of the same video at once, such as a 9:16 for TikTok and a 16:9 for YouTube. A pipeline can also compile short videos you already published into a single longer one.
Yes. Every instruction sent to a model is an editable template, and each step of a pipeline chooses its own model. You can also have a source run its items past a model before anything is made from them, to reject, score or rewrite them.
Model usage. Every step records what it consumed and is priced against the model it used, so a run with a cheaper script model costs fewer tokens. The ledger shows the charge per step.
To the channels you connect, one publish job per channel so a failure on one platform does not block the others. You can also download the file and post it yourself.