GPT-6 Sol Becomes Recommended Model in Conversation AI

TL;DR

HighLevel now defaults new prompt-based Conversation AI agents to GPT-6 Sol. Agents on deprecated GPT-5 models have been shifted automatically, while other existing agents stay on their current model.

Who this affects

HighLevel has not confirmed specific account tiers or user permissions affected by this change.

Agency owners and sub-account users managing prompt-based Conversation AI agents need to review their setups, particularly if those agents previously ran on deprecated models.

What changed

• GPT-6 Sol is now listed first in the Model menu and marked as Recommended for prompt-based Conversation AI agents.

• New prompt-based Conversation AI agents use GPT-6 Sol by default.

• Existing agents keep their previously chosen model.

• GPT-5 and GPT-5 Mini have been deprecated, and any agents running on them were automatically moved to GPT-6 Sol.

• GPT-4.1 is being retired inside Conversation AI due to OpenAI deprecation.

• Plan gating, usage limits, and rollout status have not been confirmed by HighLevel.

Where to find it

Open your sub-account and go to AI Agents → Conversation AI → Build tab → Model menu.

How agencies will use this

When deploying a new client sub-account from a snapshot, an agency can build a prompt-based Conversation AI agent without needing to adjust model settings manually. The agent will adopt GPT-6 Sol by default in the Build tab.

For existing client sub-accounts where workflows pass inbound leads to Conversation AI via automated triggers, operators should inspect agents that were previously configured on GPT-5 or GPT-5 Mini. Since HighLevel moved those agents to GPT-6 Sol automatically, agencies will want to run test conversations to ensure response behavior and prompt adherence continue as expected.

Our take

This is a standard model lifecycle update. With OpenAI retiring older models like GPT-4.1, HighLevel is cleaning house and deprecating GPT-5 variants in favor of GPT-6 Sol.

It is a minor operational change. If your agents were shifted automatically from GPT-5 or GPT-5 Mini, review your conversation logs this week to verify performance. If you are using another active model, you do not need to take action.