Why doesn't vapi use OpenAI's transcriber to solve...
# support
a
open ai's automatic multilingual transcriber instead of deepgram
v
Hi @Abed Malak, I'm an AI assistant for Vapi. I'll analyze available sources and get back to you shortly.
----------------------------- **While I generate an answer, here are some sources that might be helpful**: * [Introduction]() * [Billing Examples]() * [Web]() * [Create Phone Call]() -----------------------------
### Answer Vapi supports multiple providers for its speech-to-text transcriptions, including Deepgram and Talkscriber [(1)]() . The choice of these providers might be based on various factors such as performance, accuracy, language support, model availability, or other specific features that align with Vapi's requirements for delivering efficient and scalable voice AI solutions. While OpenAI offers powerful language models, the documentation doesn't mention the use of OpenAI's transcriber for speech-to-text purposes within Vapi's infrastructure. This decision could be influenced by the specific strengths or capabilities of the chosen providers in terms of transcription accuracy, latency, or the specific needs of Vapi's platform and its users. For detailed information about the supported providers and their capabilities, checking the documentation directly or reaching out to Vapi for more specific insights into their technology choices could provide more clarity.
s
It is slow.
You can give it a try it already there in the vapi.
a
couldn't find it here, where can I find it
s
select talkscriber then whisper
a
ok thanks!