Multi-LLM Providers in Conversation AI: Run Your Agents on Claude and Gemini
Conversation AI agents are no longer limited to OpenAI models. You can now run any agent on Anthropic Claude or Google Gemini, picked from the same Model menu as the GPT models. Choose the model that fits each agent: a…
Conversation AI agents are no longer limited to OpenAI models. You can now run any agent on Anthropic Claude or Google Gemini, picked from the same Model menu as the GPT models. Choose the model that fits each agent: a top model for complex sales conversations, or a faster, lower-cost model for simple questions and high-volume replies.
- Claude and Gemini next to GPT: The Model menu now lists Claude Opus 5.5, Claude Sonnet 5.5, Claude Haiku 5.5 and Claude Haiku 4.5, plus Gemini 3.8 Flash, Gemini 3.7 Flash and Gemini 3.5 Flash Lite.
- See the cost at a glance: Every model shows a relative cost from $ to $$$$, so you can weigh quality against cost before you switch.
- Choose a secondary model from any provider: Click the settings icon next to the model to see the Primary Model and pick a Secondary Model. The secondary model can come from any provider, but it can't be the same as the primary. When Primary model is down, the agent automatically switches to the secondary model.
- Switch models by chatting: In Build with AI, ask for the model you want, for example "Switch the model to Claude", and the agent's primary model is updated for you.
- One model per agent: Each agent has its own model, so different agents in the same sub-account can run on different providers.
- Nothing else changes: Your prompt, actions, knowledge base, channels and settings stay as they are when you switch models.
- Enable Conversations AI Provider Support under Settings → Labs.
- In AI Agents → Conversation AI, open your agent.
- Open the Model menu and choose a Claude or Gemini model.
- Optionally, click the settings icon next to the model and choose a Secondary Model.
- Try the agent in Test your agent, then click Save.
- The right model for each job: Use a top model such as Claude Opus 5.5 where replies need the most care, and a lighter one where speed matters more.
- Control your costs: Lower-cost models such as Claude Haiku 5.5 and Gemini 3.5 Flash Lite suit simple, high-volume conversations.
- Less reliance on one provider: Pick a secondary model from a different provider than your primary.
- Test after you switch: Each model writes in its own way, so test your agent with the new model before you save.
- GPT 5 and GPT 5 Mini retire on 30 Oct: Agents on these models should be moved to another model before then.
Multi-LLM provider support is available in Labs. Enable it under Settings → Labs → Conversations AI | Claude & Gemini Model Support (Location).