HighLevel Launches OpenAI GPT Live Speech-to-Speech
TL;DR
HighLevel is a launch partner for OpenAI's GPT Live speech-to-speech model, now available within Voice AI.

TL;DR
HighLevel launched OpenAI's GPT Live speech-to-speech model. This new model processes speech simultaneously, allowing for natural conversations.
Every call using this model is free until September 17th.
Who this affects
HighLevel has not confirmed who is affected by this update.
What changed
HighLevel is a launch partner for OpenAI's GPT Live speech-to-speech model.
This new model listens and speaks simultaneously.
It differs from turn-based systems, allowing for natural interruptions in conversations.
The model delegates tasks like booking lookups or workflow actions to the backend while maintaining conversation flow.
Every call on this model is free until September 17th.
HighLevel has not confirmed any limits or plan gating for this feature.
HighLevel has not confirmed the rollout status.
Where to find it
This feature is available within Voice AI.
HighLevel has not published the exact navigation path yet, but it will live within the Voice AI settings or configuration options.
How agencies will use this
Agencies can configure this new speech-to-speech model for their clients' automated communications. For instance, a real estate agency can set up an AI agent to handle inbound calls from potential buyers. When a prospect calls, the AI agent, powered by GPT Live, can engage in a fluid, human-like conversation. If a prospect asks about available showing times for a specific property, the AI can delegate a workflow action to look up the agent's calendar, retrieve availability, and communicate those times back, all without pausing the conversation for a backend lookup. The prospect can interrupt to ask follow-up questions about a particular time slot, and the AI will adapt in real-time, improving the caller experience significantly compared to previous turn-based systems.
Our take
This is a major update. The shift from turn-based to simultaneous speech processing is a significant leap for automated voice communication. The ability for natural interruptions and simultaneous processing means the AI will sound less robotic and more human, which should improve engagement and conversion rates for clients. The free calls until September 17th offer a window to test and implement this. HighLevel has not confirmed who is affected, nor has it provided details on rollout or specific plan gating. HighLevel has not confirmed the full significance reason beyond its major impact.
Source: Official HighLevel announcement


