We’re excited to unveil our brand-new voice AI model today, combining the conversational elements you already love with a new benchmark for low-latency speech-to-speech technology.

Speed and accuracy make-or-break customer satisfaction, and even the slightest difference or delay can cost your business in reputation and potential revenue. Our new model takes a huge leap forward in instruction following and tool calling reliability, meaning you can rest assured that your agents will deliver the results you and your customers want. 

A World Leader in Speech Intelligence

We have solidified our position as a top-performing architecture in the latest VoiceBench evaluations, where our model secured a place among the highest-ranked engines globally for speech intelligence. This independent verification offers telecoms providers a significant strategic advantage, providing a high-performance solution that sits among the best, while maintaining a focus on technical excellence. By integrating Apollo, you are adopting one of the most capable and rigorously tested speech architectures available today – a move designed to drive higher CSAT scores, accelerate your lead generation pipeline, and position your brand at the absolute forefront of industry innovation. 

Instant Responses for Natural Conversations

In voice communication, a one-second delay makes a conversation feel mechanical and disjointed. Our new model introduces a 20% latency reduction, bringing response times down to a blistering 0.8 seconds. This speed is vital for natural turn-taking, preventing the awkward overlaps that frustrate callers. For your customers, this creates a seamless experience that feels like a fluid human interaction, directly driving higher satisfaction and brand loyalty. As a reseller, you can be confident in your AI agent's ability to deliver customer service to the same standard as your human staff. 

Five Times the Power for Complex Queries

The move to 355 billion parameters represents a massive 5x increase in “brain power” over our previous model. This expanded capacity allows our agents to navigate complex, multi-step queries and sophisticated tool-calling with ease. For channel partners, this transforms the AI agent from a simple greeting bot into a fully capable digital employee, expanding your team without the additional overheads a regular employee requires. It can now resolve high-level issues – like database cross-referencing or intricate scheduling – on the very first call.