### Answer When your Conversation Flow agent stops responding during node transitions, here are troubleshooting steps based on documented solutions: Check Your Transition Conditions Review your transition conditions to ensure they precisely match your intended triggers [(1) ]() . The transition may be failing if conditions aren't being met, even when it seems like they should be [(2) ]() . Verify Function Schema If transitions occur after function/tool calls, check your function schema [(3) ]() . Issues with function schemas can cause the agent to become silent after execution [(4) ]() [(5) ]() . Split Complex Nodes If a node contains too many instructions, split it into multiple nodes [(1) ]() . This improves consistency and reduces the chance of the agent getting stuck. Add Transition Examples Provide transition finetune examples to help the model understand your expectations . This can improve transition accuracy. Make Conditions More Flexible Consider making your transition conditions more flexible and general , especially if users might respond in unexpected ways. Adjust the LLM Model Try using a more capable LLM model for problematic nodes . If you're using Gemini 3.0 Flash and experiencing silence after tool calls, switching to Gemini 2.5 Flash may help [(5) ]() . **Revie

w Call Logs** Check call logs for speech recognition errors, timing issues, or overlapping speech that might prevent transitions [(2) ]() . If the issue persists after trying these solutions, you may want to reach out to support with specific call IDs for further investigation.

​ To learn more, see the sources I considered: [[1] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) [[2] conversation flow doesn't transition correctly](https://discord.com/channels/1186424376300679209/1342666714248118343) [[3] Conversation flow issue](https://discord.com/channels/1186424376300679209/1347631304790900736) [[4] Conversation Flow agent potentially stay silent after function node](https://community.retellai.com/t/conversation-flow-agent-potentially-stay-silent-after-function-node/135) [[5] Gemini 3.0 Flash Agent stay silent after tool function completion](https://community.retellai.com/t/gemini-3-0-flash-agent-stay-silent-after-tool-function-completion/336) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer I understand the frustration when support channels aren't responsive. Here are the official support options available for Retell AI: Documentation Resources Visit the [Documentation]() for comprehensive guides and troubleshooting steps [(1) ]() . Certified Partners You can also visit [Hire certified Retell partner]() for professional assistance with your implementation . While waiting for support, based on the previous troubleshooting steps shared, the most common causes for agents stopping during node transitions are: 1. Transition condition issues - ensure conditions precisely match your intended triggers 2. Function-related silence - if using function nodes, there have been documented incidents where agents stayed silent after function execution [(2) ]() [(3) ]() [(4) ]() [(5) ]() 3. Node complexity - splitting complex nodes into multiple simpler nodes can improve consistency [(1) ]() 4. LLM model selection - switching to a more capable or different model version may resolve the issue [(1) ]() [(4) ]() If you can share a specific call ID where this happens, that would help diagnose the exact cause of the silence during transitio

ns.

​ To learn more, see the sources I considered: [[1] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) [[2] Conversation Flow agent potentially stay silent after function node](https://community.retellai.com/t/conversation-flow-agent-potentially-stay-silent-after-function-node/135) [[3] Conversation flow agent stops speaking after calling function](https://discord.com/channels/1186424376300679209/1394277381782638672) [[4] Gemini 3.0 Flash Agent stay silent after tool function completion](https://community.retellai.com/t/gemini-3-0-flash-agent-stay-silent-after-tool-function-completion/336) [[5] Function Calling not working in Conversation Flow](https://discord.com/channels/1186424376300679209/1394311821388681227) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer In rigid mode, the agent only processes the active node's prompt rather than compiling all nodes into a single prompt [(1) ]() . This is fundamentally different from flex mode, where all node instructions, transitions, and tool descriptions are compiled into one prompt, which can significantly increase token count and LLM costs when flows exceed 3,500 tokens . Since rigid mode sends only the current node's context to the LLM, having 250+ nodes shouldn't inherently overwhelm the system from a prompt-size perspective. The agent processes nodes sequentially, not all at once . However, the complexity of your workflow could still contribute to issues: Potential factors with large rigid-mode flows: - Transition logic complexity - With 250+ nodes, there are many potential transition paths that need to be evaluated - State management - The system needs to track which node is active and evaluate transition conditions - Network/processing overhead - More nodes mean more potential points where timing or execution issues could occur The documented silence issues after function nodes [(2) ]() and transition problems don't specifically mention node count as a factor. The sources focus more on: - Function schema issues - Transition condition precision - Node instruction complexity - LLM model selection If you suspect the large workflow is causing problems, consider breaking it into smaller components or using modular conversation flows where appropriate. This aligns with the recommendation to split complex nodes for better consistency [(1) ]() . Without specific documentation on node count limits for rigid mode, I'd recommend t

esting with a simplified version of your flow to isolate whether the 250+ node count is the root cause.

​ To learn more, see the sources I considered: [[1] Flex Mode](https://docs.retellai.com/build/conversation-flow/flex-mode) [[2] Conversation Flow agent potentially stay silent after function node](https://community.retellai.com/t/conversation-flow-agent-potentially-stay-silent-after-function-node/135) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.