Manish
03/01/2025, 10:08 AMheuristic
)
07:45:08:801
[CHECKPOINT]
Model request started
07:45:08:802
[LOG]
Model request started (attempt #1, google/gemini-flash-1.5, openrouter)
07:45:09:496
[CHECKPOINT]
Model sent start token
07:45:09:501
[WARN]
Model request failed (attempt #1, reason: (toolType not set))Vapi Support Bot
03/01/2025, 10:09 AMtype
property. This should be one of the predefined types required by the API, such as dtmf
, endCall
, transferCall
, etc. Make sure there are no missing or erroneous type
fields.
2. **Review Recent Changes**: If there have been any recent changes to your system configuration or code, review them to see if they might have impacted the tool calls.
3. **Verify Completion Message Configurations**: Check if the configuration of the completion messages is set up correctly. The issue can stem from completion messages not being set up appropriately for specific tool calls.
4. **Inspect JSON Payload**: Ensure that the JSON payload being sent is correctly formatted and that required fields are not omitted. VAPI expects certain fields and structures, and missing these can result in errors.
5. **Check for Logic Verification**: Make sure that the assistant's logic for providing feedback after a tool call is intact and hasn't been unintentionally altered.
6. **Review Webhook and Server Issues**: Confirm that any webhooks associated with VAPI are correctly configured and matched with their corresponding endpoints. This includes checking URL correctness and network accessibility.
If these steps do not resolve the issue, it may be helpful to reach out to VAPI support with the details of your error and the CallId you provided for further assistance.
Let me know if you need any more help! 😊
Sources:Vapi Support Bot
03/01/2025, 10:09 AMManish
03/01/2025, 10:12 AMKyle Brunker
03/03/2025, 1:06 PMManish
03/15/2025, 11:50 AMManish
03/17/2025, 1:00 PMShubham Bajaj
03/17/2025, 3:35 PM