I have been consistently using a Make scenario with the AssemblyAI module to convert my class audio recordings into text. I believe I originally created this scenario around September–October 2025.
However, sometime around November or December, Make introduced a new AssemblyAI module. Since then, whenever I try to recreate the same scenario using the new AssemblyAI module, I keep encountering errors. First, it says that the target language is not set. When I set the language, it then throws another error saying that the format settings are not configured.
The core problem is this:
Even if I duplicate an old scenario that I created in the past and that is still working perfectly, the newly added AssemblyAI module inside the copied scenario does not work.
The original scenario continues to run without any issues, even though it does not have any explicit target language or format settings configured. But any AssemblyAI module that is newly created after a certain point in time simply fails to transcribe audio into text at all.
I have already asked ChatGPT and Gemini for help, but neither was able to solve the issue. That is why I suspect that something may have broken when AssemblyAI updated their module at a specific point in time.
To be clear again:
Existing scenarios that already contain the AssemblyAI module work normally.
Hello @scsvy1 , welcome to make.com community, I have worked and have experience with Make.com and l will love to collaborate with you on this you can schedule a call Here and you can checkout my upwork profile Here, for my pastworks and certifications
Hey @scsvy1 , This isn’t your setup, it’s a breaking change in the new AssemblyAI Make module. Older scenarios are running on legacy defaults, while newly created modules require explicit language and format fields and still fail due to a schema/version mismatch. Duplicating a scenario rebinds it to the new module, which is why it breaks.
Workaround: bypass the native module and call AssemblyAI via the Make HTTP module with all parameters set explicitly. This is a module regression, not user error.
This looks like a breaking change in the new AssemblyAI module on Make. Old scenarios still rely on legacy defaults that no longer apply to newly created modules. I can help you isolate exactly what changed and provide a working workaround—or confirm if this needs a Make support escalation. Want me to take a quick look at the scenario setup with you?
This looks like a breaking change in the newer AssemblyAI module rather than an issue with your scenario. Older scenarios keep working because they are effectively locked to the legacy module version, while any newly created AssemblyAI module enforces new required fields around language and format that were not previously required.
I’ve seen this exact pattern in Make where duplicated scenarios still behave like legacy ones, but newly added modules fail due to schema changes.
I work mainly on production grade Make.com automations and have handled similar transcription and media pipeline issues before. You can check my website portfolio for related examples.
A reliable workaround here is either explicitly mapping the new required fields end to end or bypassing the native module entirely and calling AssemblyAI via the HTTP module for full control.
Happy to help further. You can book a quick call here with me or continue the discussion privately in my Fiverr workspace. Whatever works for you best.