some calls are working some are not this is very concerning, we have a demo to client soon. Please team @Vapi look into this

Apologies for the delay.

🟡 13:42:52:185 Couldn't Complete Say Request { "task": { "content": "Hi Sahil, I am Spencer, an Network Engineer - Routing & Switching designed and developed by NTT DATA. I am really excited about this opportunity and look forward to sharing how I can contribute to your team", "endCallOnError": true, "interruptible": false }, "error": {} } 🟡 13:40:50:595 Couldn't Complete Say Request { "task": { "content": "Hi Sahil, I am Arya, an AWS Engineer designed and developed by NTT DATA. I am really excited about this opportunity and look forward to sharing how I can contribute to your team", "endCallOnError": true, "interruptible": false }, "error": {} } The calls are ending with as the customer ended the call because you have set end call on error flag to true for the same messages.

same --> say \*

hey @Shubham Bajaj but i am not setting this paramenter to true anywhere in my config then is it the default value ?

{ name: bot?.name, firstMessage:

Hi ${user?.firstName}, I am ${bot?.botName}, an ${bot?.roleId?.name} designed and developed by ${currentTenantName}. I am really excited about this opportunity and look forward to sharing how I can contribute to your team

, transcriber: { provider: "deepgram", model: "nova-3", language: "en-US", }, voice: { provider: "playht", voiceId: bot.voice, }, model: { provider: "openai", model: "gpt-4.1-mini", messages: [ { role: "system", content:

Current Date and Time is ${currentDateTime} (${localTimeZone})\n\n${voicePrompt}

}, { role: "assistant", content:

User Details: I am interacting with ${user.firstName} ${user.lastName} whose email address is (${user.email}), who holds the role of ${user.roleId?.name || "unspecified role"}.

} ], tools: [ { type: "function", messages: [], function: { name: "web_search", parameters: { type: "object", properties: { search_query: { type: "string" } }, required: ["search_query"] }, description: "Search the web using a query string and return relevant results." }, async: false, server: { url:

${baseURL}/${user.id}/webSearch

, timeoutSeconds: 300 } } ], }, backgroundDenoisingEnabled: true, modelOutputInMessagesEnabled: true, startSpeakingPlan: { waitSeconds: 2 }, stopSpeakingPlan: { numWords: 0, voiceSeconds: 0, backoffSeconds: 2 }, analysisPlan: { successEvaluationPlan: { enabled: false }, summaryPlan: { enabled: false } }, maxDurationSeconds: 1800 }

also one more thing i would like to know if i change the voice provider from playht to vapi will there be any cost difference ?

Your SAY request failed because the voice provider returned a text-to-speech error, which caused the call to end. To handle this gracefully, you can configure a fallback Voice. Alternatively, try using VAPI Voices. Also, the end call flag is something which is enabled by default to tackle these kinds of issues.