m
mata
05/07/2026, 12:03 PM@AI Evy Hey! 👋
I'm building an automated testing pipeline for my Retell voice agent using n8n.
My current flow generates a simulated user prompt (with identity, goal, personality)
and I manually paste it into the AI simulated chat in the Retell UI.
Is there any API endpoint that allows me to:
1. Submit a simulated user system prompt programmatically
2. Trigger a test conversation against a specific agent
3. Get back the transcript / node transitions / tool calls as a result
Basically I want to automate what the "AI simulated chat" feature does in the UI —
without having to copy-paste the prompt manually each time.
Is this possible via the Retell API, or is there a recommended approach for
automated end-to-end testing of voice agents?
Thanks!
a
AI Evy
05/07/2026, 12:03 PMThanks for tagging me @mata. I'll look through relevant sources and get back to you shortly.
AI Evy
05/07/2026, 12:03 PM### Answer
Based on the available documentation, API access for programmatic simulation testing is not currently available [(1) ]() .
The AI Simulated Chat feature you're using in the dashboard does not have corresponding API endpoints for:
- Submitting simulated user prompts programmatically
- Triggering test conversations via API
- Retrieving simulation test results programmatically
Current Testing Options:
The documentation shows two testing approaches available through the UI [(2) ]() :
1. Manual Chat Testing - Interactive testing in the LLM Playground [(3) ]()
2. AI Simulated Chat - Where you define user scenarios with identity, goal, and personality [(2) ]()
For the AI Simulated Chat, you can save test cases and run batch simulations, but these must be triggered through the dashboard [(2) ]() [(4) ]() .
Community Feedback:
There have been requests for API access to simulated test results and test definitions, but there are currently no updates on when this functionality will be available [(1) ]() .
Potential Workaround:
While not a direct replacement for simulation testing, you could consider using the regular call APIs to create actual test calls programmatically, though this would incur call costs and wouldn't provide the same simulation-specific features like the AI-generated user responses.
I recommend checking the [Retell AI changelog](
AI Evy
05/07/2026, 12:03 PMellai.com/changelog) periodically for updates on API testing capabilities, as this appears to be a feature that's been requested by the community .
AI Evy
05/07/2026, 12:03 PM
To learn more, see the sources I considered:
[[1] Api access to simulated test results + test definitions](https://community.retellai.com/t/api-access-to-simulated-test-results-test-definitions/270)
[[2] Automatically test your agent](https://docs.retellai.com/test/llm-simulation-testing)
[[3] Manually test your agent](https://docs.retellai.com/test/llm-playground)
[[4] Retell AI Introduces Simulation and Batch Testing for AI Agents | Retell AI](https://www.retellai.com/blog/retell-ai-introduces-simulation-and-batch-testing-for-ai-agents)
AI Evy
05/07/2026, 12:03 PMFeel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
AI Evy
05/07/2026, 12:04 PMThanks for the feedback, glad I was helpful.