Skip to main content

​
What Is Dual Chat?

Dual Chat lets you send a single message to two different AI models at the same time and see their responses side by side. This is useful when you want to compare how different models handle the same question, verify an answer across providers, or find the best model for a particular task. Both models receive your exact message simultaneously and stream their responses in parallel. You see two columns, each showing one model’s output in real time.

Left Panel

The first AI model’s streaming response.

Right Panel

The second AI model’s streaming response.

​
Starting a Dual Chat

1

Open a new Dual Chat

Click Dual Chat in the sidebar navigation. This opens the dual chat interface with two model selectors at the top.
2

Select your first model

Use the left model selector to choose the first AI model. You can pick any model from any supported provider — OpenAI, Anthropic, Google, or xAI.
3

Select your second model

Use the right model selector to choose the second AI model. Pick a different model or even the same model from a different provider to compare implementations.
4

Send your message

Type your message in the single input field at the bottom. When you press Enter, RelayHub sends the message to both models simultaneously.
5

Compare responses

Watch both responses stream in side by side. Scroll either panel independently to review longer outputs.

​
When to Use Dual Chat

Trying to decide which AI model works best for your use case? Run the same prompts through two models and compare quality, tone, and accuracy. This is especially helpful when your organization is choosing a primary model for a specific workflow.

​
Working with Dual Chat Results

You can continue the conversation in Dual Chat by sending follow-up messages. Each model maintains its own context from prior messages in the session, so follow-ups build on each model’s previous answers independently.
Use Dual Chat at the start of a project to pick your model, then switch to standard chat for the rest of the work. This saves you from switching models mid-conversation later.

​
Practical Tips

  • Mix providers for best comparison. Comparing an OpenAI model against an Anthropic model gives you more diversity than comparing two OpenAI models.
  • Test with representative prompts. Use real examples from your actual work rather than generic test questions. Model performance varies significantly by domain.
  • Check both reasoning and format. One model might give a better answer but in a worse format, or vice versa. Consider both when evaluating.
Dual Chat uses tokens from both models for each message. Your usage reflects the combined cost of both models. Keep this in mind when running extended comparison sessions.

​
Limitations

Dual Chat is designed for model comparison. Some features available in standard chat — such as sharing and real-time collaboration — are available only on standard conversations. If you find a model you prefer during a Dual Chat session, start a new standard chat with that model for your ongoing work.