Speaker identification/separation is not working for some iOS transcripts

What to do

• Check whether your product depends on Granola. Check your stack

  Paste a package.json, lockfile, Terraform or a list of domains at down.guide/stack to see every provider you depend on.

• Confirm your retries, queues and scheduled jobs caught up after the 3h 46m it lasted.

  Anything that failed between 2026-04-08 16:59 UTC and 2026-04-08 20:45 UTC may need a replay.

• Subscribe to Granola's feed to hear about the next one first. RSS feed

  https://down.guide/feed?provider=granola works in Slack (/feed subscribe), Discord feed bots and any RSS reader.

Updates from Granola 4

1. investigatingApr 8, 2026, 4:59 PM UTC

   Usually, our iOS app separates the transcript into Speaker A, Speaker B - this seems to be failing for some users at the moment. We're currently investigating and will provide an update here shortly.

   provider

2. identifiedApr 8, 2026, 5:58 PM UTC

   We've identified the cause of this issue and we're implementing a fix now.

   provider

3. monitoringApr 8, 2026, 7:49 PM UTC

   We've deployed a fix for this and are monitoring diagnostics to confirm that it's resolved. Apologies for any disruption - get in touch at [hey@granola.so](mailto:hey@granola.so "hey@granola.so") if you're still having problems!

   provider

4. resolvedApr 8, 2026, 8:45 PM UTC

   We're still evaluating the impact here and may be able to backfill speaker diarization for existing affected transcripts over the next couple of days, but going forwards this issue is resolved!

   provider

Corrections

None so far.

Something wrong on this page? Report an error or send us a note. Every fix is listed here.