Everything we've shipped, newest first — with links to the guides that explain it.
Spotted something broken, or want something we don't have? Report it from your workspace sidebar and we'll email you when it's done.
When a YouTube post failed because the day's upload allowance had run out, the message told you YouTube was having a known issue and was investigating. That was not accurate, and it left you waiting for something to be fixed at YouTube's end that was never going to fix itself.
You now see what actually happened. YouTube limits how many videos an app like Autoposter can upload each day, and that allowance is shared by everyone posting to YouTube through Autoposter — it is not a limit on your channel. The message now says so, and tells you when the allowance resets (midnight US Pacific Time) so you can reschedule the post.
We have applied to YouTube for a much larger daily allowance. Until it is granted, scheduling YouTube posts earlier in the day gives them the best chance of going out.
Some videos — iPhone recordings and anything not already in the standard format — have to be converted before they can be posted. Until now that conversion happened while you waited: the upload bar reached the end and then sat there, sometimes for minutes, before the file appeared in your library. Worse, when several people uploaded at once those conversions piled up and slowed the whole app down, which is what was behind the calendar going blank and then coming back.
Conversion now happens in the background, a few seconds after your upload lands. The file appears in your library straight away and you can carry on working while it finishes.
The only thing to know: a video that is still converting cannot be posted for those few seconds. If you upload a video and immediately publish it, we hold the post until the video is ready rather than posting the unconverted file — so it may go out a minute or two later than you clicked. Quality is unchanged.
Uploading to the media library has been slow and unreliable for some of you — files crawling along, sitting at the same percentage for minutes, and sometimes failing outright after a long wait and needing to be started again. If your internet connection is shared, busy in the evenings, or simply not very fast, this was hitting you hardest.
Large files now upload in smaller pieces, and one piece at a time instead of three at once. Sending three pieces together meant they competed with each other for the same connection, so on a busy line all three crawled and the upload was far more likely to time out. One piece at a time gets the whole connection to itself, so each piece finishes quickly and the upload keeps moving.
You do not need to change anything, and nothing about the files themselves has changed — same formats, same size limits, same quality. If an upload is still giving you trouble, the troubleshooting guide below covers what to check.
Some uploads were coming back with every platform marked red and a vague note about the content being flagged — no detail about what the problem actually was, and nothing you could act on. That was never a real verdict on your content. It meant the compliance checker had not managed to read the file at all, and we had nothing useful to show you.
The compliance checker has been upgraded and now analyses this content properly. Around 9 in 10 of the affected uploads come back with a normal per-platform rating and specific, actionable recommendations. We have re-checked everything that was stuck in this state, so your existing media library is updated too — nothing to re-upload.
A small number of files still cannot be read by the checker. Those are held back for you to review yourself rather than being waved through — but they now say so plainly, instead of reading like a judgement on your content.
There is now a Report an issue link at the bottom of your sidebar. Use it to report anything broken or to suggest a feature — attach a screenshot or a short screen recording if it helps explain.
Everything you send lands in a queue our team works through. You get an email when it arrives, and another when it is done — or when we decide not to build it, with the reason why. You can see the status of everything your workspace has raised on the same page.
This changelog is the other half of that loop: when a fix ships, the email links you straight to the entry here.
Some connected accounts were showing a blank profile picture in the sidebar and account list. When an account was connected, the picture was occasionally failing to save to our storage while still being recorded as saved — so nothing looked wrong until you noticed the empty circle.
We have restored the pictures for 92 affected accounts and fixed the underlying cause, so new connections verify the image really saved before recording it. A weekly check now repairs any that slip through on their own.
A small number of accounts could not be restored because their connection to the platform has expired — reconnecting those accounts will bring the picture back.
Your dashboard now shows median views next to average views on every platform that reports views. One viral post can drag an average far above what your posts actually do — the median tells you what a typical post gets, so the gap between the two numbers tells you whether a period was carried by a single hit.
Instagram accounts also get average watch time across the reels in the period, with the number of reels it was measured over.
When a post went out to some accounts but failed on others, the whole post was recorded as failed — so its images and videos showed in the media library as though they had never been published. The library now reads the per-account result, so anything that actually went live is labelled as published.
Clean, high-bitrate videos were being re-compressed before upload, which is why some reels looked softer on Instagram than in your editor. Videos that already meet Instagram's specification now pass through untouched, and only genuinely incompatible files (10-bit or HDR sources) are converted.
Meta retired a family of Facebook metrics in June, which stopped our whole insights request for Facebook Pages — leaving a blank spell in your Facebook figures. We have moved to Meta's replacement metrics, which backfill 90 days, and fixed a separate mismatch in how Facebook post IDs were matched that had been dropping most Facebook posts out of tag and video reporting.
Threads accounts were dropping out at exactly 60 days and needing a manual reconnect, because their tokens were being skipped by our refresh routine. They now refresh like every other platform.
A read-only API for pulling your own analytics — account summaries, trends, top posts and follower history — into your own dashboards and reports. Tokens are issued per workspace with scoped permissions.
Group your uploads into folders, filter the library by folder, and assign folders in bulk. The media picker in the post composer respects them too, so finding the right clip in a library of thousands is no longer a scroll.