GPT-6 Sol is now the recommended model for prompt-based Conversation AI agents in EveryCatch. Conversation AI has been tuned to work especially well with GPT-6 Sol, so your agents can give more accurate answers than with GPT-4.1, while keeping the cost per reply at a similar level.
Key capabilities
- Recommended in the Model menu: GPT-6 Sol now appears first in the Model menu for prompt-based agents and is marked as Recommended, so you can find it quickly.
- New agents start on GPT-6 Sol: Any new prompt-based Conversation AI agent you create will use GPT-6 Sol by default, so you start with our best tested option immediately.
- Your choice stays: Existing agents keep using whichever model you already selected, unless you decide to change them to GPT-6 Sol yourself.
- More accurate at the same cost: GPT-6 Sol reaches a 98.4% pass rate on our Conversation AI evaluation, compared with 96.3% on GPT-4.1, while keeping the cost per reply at a similar level.
- Fast replies: Typical reply generation time is about 1.68 seconds, with the slowest 5% of replies arriving in about 2.19 seconds.
Changes to older models
- GPT-5 and GPT-5 Mini are deprecated: Agents that were using these models now automatically run on GPT-6 Sol, with reply generation up to 4.5 times faster. You do not need to change any settings for this update to apply.
- GPT-4.1 is being retired: OpenAI is deprecating GPT-4.1, and it will be removed from Conversation AI. Update any agents that still use GPT-4.1 to GPT-6 Sol so they continue working without interruption.
How to get started
- Go to AI Agents → Conversation AI and open your agent.
- Open the Model menu on the Build tab.
- Select GPT-6 Sol (marked Recommended).
- Click Save, then try it in Test your agent to see the new behavior.
Why it matters
- Better answers out of the box: Every new prompt-based Conversation AI agent starts on the most accurate model we have evaluated, so you begin with stronger responses from day one.
- No extra cost: You get higher accuracy than GPT-4.1 while keeping the cost per reply at about the same level.
- Faster replies: Agents that were using GPT-5 and GPT-5 Mini now respond up to 4.5 times faster on GPT-6 Sol, which helps conversations feel smoother to your contacts.
- An easy move off GPT-4.1: Updating agents from GPT-4.1 to GPT-6 Sol is a simple model change in the Build tab before GPT-4.1 is fully retired.
Important notes
- Test after you switch: Each model has its own response style and behavior, so use Test your agent after changing to GPT-6 Sol to confirm that replies match your tone and workflow before you rely on it in production.
- Prompt-based agents: This recommendation applies to prompt-based Conversation AI agents. Other types of agents are not affected by this change.
