LocalAI version:
latest docker, cloned yestarday
Environment, CPU architecture, OS, and Version:
Linux denis-pc 7.0.3-arch1-2 #1 SMP PREEMPT_DYNAMIC Fri, 01 May 2026 15:49:22 +0000 x86_64 GNU/Linux
Describe the bug
When creating a new speech-to-speech pipeline (icon of a phone) I cannot select parakeet in the speech-to-text field, only in LLM, but parakeet is not a LLM and nvidia clearly says it's a speech-to-text: https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3
To Reproduce
Add the nvidia/parakeet-tdt-0.6b-v3 from huggingface, you will only be able to use it as an LLM (ofc it's not going to work)
Expected behavior
For speech-to-text to be usable as speech-to-text and not as an LLM
Logs
No need for logs
Additional context
No need for additional context
LocalAI version:
latest docker, cloned yestarday
Environment, CPU architecture, OS, and Version:
Describe the bug
When creating a new speech-to-speech pipeline (icon of a phone) I cannot select parakeet in the speech-to-text field, only in LLM, but parakeet is not a LLM and nvidia clearly says it's a speech-to-text: https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3
To Reproduce
Add the nvidia/parakeet-tdt-0.6b-v3 from huggingface, you will only be able to use it as an LLM (ofc it's not going to work)
Expected behavior
For speech-to-text to be usable as speech-to-text and not as an LLM
Logs
No need for logs
Additional context
No need for additional context