Can you give me the price and benefits that come with the enterprise plan?

@Dwayne Certainly! The Vapi Enterprise Plan includes a suite of benefits designed to meet the needs of larger organizations and those requiring advanced features. Here are the key benefits: - Guaranteed low latency on our Enterprise cluster for optimal performance. - Hands-on 24/7 support to assist you whenever you need it. - A shared Slack channel with our team for direct communication and quick resolution of queries. - Included minutes with volume pricing, providing cost-effective solutions as your usage grows. - Regular calls with our engineering team 2-3 times per week to ensure your needs are being met and to provide personalized assistance. - Reserved concurrency and higher rate limits to handle your high-demand applications seamlessly. - Access to the Vapi SIP trunk for telephony, allowing for robust voice communication capabilities. For the pricing details specific to your organization's needs, including volume pricing and any additional services, please fill out the form on our "Contact us" page so we can provide you with a personalized quote. Would you like assistance with filling out the form or have any other questions about the Enterprise Plan?

Vapi Support Bot

@Dwayne I apologize for any confusion. For a team with just 2 employees, the pricing for the Enterprise Plan would still depend on the volume of calls you expect to handle and the number of different assistants and phone numbers you plan to create with Twilio. Since you're looking to create multiple assistants and use different phone numbers, here's a basic cost breakdown based on the voice pipeline components you might use: - Deepgram: ≈ $0.01/min for speech recognition. - Custom Model: ≈ $0.04/min, but this can vary widely depending on the complexity and specific requirements of your custom model. - PlayHT: ≈ $0.07/min for text-to-speech services. These costs will be multiplied by your call volume to determine the overall price. Additionally, Twilio's pricing for phone numbers and call handling would be separate and based on their current rates. For a precise quote tailored to your specific use case, including the number of assistants and phone numbers you need, it would be best to discuss your requirements directly with our sales team. They can provide a detailed cost estimate and explain how the Enterprise Plan can be customized for your small team. Would you like to proceed with contacting our sales team for a personalized quote?

Ah..ok I understand. What speech service and what setting offers optimum realistic human voice/speech?

@Dwayne For achieving the most realistic human-like voice and speech quality, you would want to use a state-of-the-art text-to-speech (TTS) service. Vapi integrates with several leading TTS providers, and among them, services like Google WaveNet and Amazon Polly are known for their high-quality, natural-sounding voices. When configuring your TTS service for optimal realism, consider the following settings: 1. **Voice Selection**: Choose a voice that sounds most natural and fits your application's context. Both WaveNet and Polly offer a variety of voices in different languages and accents. 2. **Speaking Style**: Some TTS services offer speaking styles that can be adjusted to sound more conversational, newscaster-like, or suited for different contexts. 3. **Speech Speed and Pitch**: Adjust the speed and pitch to match natural human speech patterns. Avoid extremes that can make the voice sound unnatural. 4. **SSML (Speech Synthesis Markup Language)**: Use SSML tags to add pauses, emphasize words, control intonation, and simulate natural speech nuances. 5. **Custom Pronunciations**: Customize the pronunciation of specific words or phrases to ensure they sound correct. 6. **Dynamic Range Compression**: This audio processing technique can make the speech clearer and more consistent in volume, contributing to a more natural sound. For the best results, it's recommended to experiment with these settings and conduct user testing to find the most natural-sounding configuration for your specific use case. Would you like more detailed guidance on setting up a TTS service or assistance with integrating it into your application?

Cartesia has been my favorite, or playht but their latency and cost is fairly high

@Dwayne Please DM me for Enterprise related queries.

Thanks @Mason that is helpful. Also I will reach out to support on Calendly.

Better than elevenlabs?

Significantly

Really? I haven’t heard that. What about it seems better? I haven’t really tried it

Elevenlabs has chunking errors with responses, Cartesia is also a different architecture it’s built with state space models which is different from pretty much everyone.

I see. What about vs play HT pricing acids which is the best model? I think I’ll try out Cartesian

I like playht but they do weird stuff with their agents some weird middle layer stuff I believe. If they fixed that and lowered the price more I’d use them

What about LLM? Are you using 4o?

@User 4o-mini