Different billing rate for TTS.
# support
t
Is there a reason why the TTS for 11Labs is different on every call? Sometimes it's 0.10/min, sometimes it's 0.15/min. But in the setup of the assistant it says it's 0.04/min.
v
Hi @Terry Ruch, 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**: * [Billing Examples]() * [Inbound Calling]() * [Dashboard]() * [Web Calling]() -----------------------------
### Answer The difference in billing rates for TTS with 11Labs you're observing could be influenced by several factors. The initial rate mentioned (0.04/min) might reflect a base or estimated cost. Variable rates per call can occur due to factors like increased usage, demand spikes, or specific features used in the calls that may incur additional costs. Because TTS pricing can vary based on the specific parameters and usage, it's often averaged or estimated in general documentation [(1)]() . For a more accurate picture of why you're seeing these variations, it would be beneficial to review the detailed billing breakdown for each call on your dashboard, where you can see the costs attributed to TTS and other services for individual calls. If you continue to see discrepancies that seem inconsistent with your service usage, it may help to directly check with the TTS provider or the Vapi support for detailed clarification on the billing structure.
s
It varies because the number of input characters changes. During one minute, the output of the LLM model can be quite large, and sometimes it can be pretty low. So, that value changes as the amount of characters also changes
t
Oh, so it really isn't per minute, but per character? Is that true for all the TTS?
s
Yup.