From 5be3559d63baf80940f0ff1f546aca893965a9f8 Mon Sep 17 00:00:00 2001 From: Steve Smtih Date: Mon, 19 Oct 2026 05:11:49 -0700 Subject: [PATCH 1/2] Document GPT-Live compatibility for assistant-based warm transfer --- fern/calls/assistant-based-warm-transfer.mdx | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fern/calls/assistant-based-warm-transfer.mdx b/fern/calls/assistant-based-warm-transfer.mdx index ae8c4ddfd..98cdfd5b1 100644 --- a/fern/calls/assistant-based-warm-transfer.mdx +++ b/fern/calls/assistant-based-warm-transfer.mdx @@ -9,6 +9,10 @@ Assistant-based warm transfer places the customer on hold while a transfer assis Use this mode when the destination must accept the call or receive context before speaking with the customer. For a direct transfer without an operator conversation, use the [Transfer Call tool](/tools/transfer-call) with its default blind-transfer mode. + +GPT-Live assistants cannot use assistant-based warm transfer. GPT-Live supports only blind transfers, and only on Twilio and Vapi SIP calls. See [GPT-Live compatibility](/gpt-live/configuration#compatibility) for the full list of supported and unsupported features. + + ## Prerequisites Before you configure the transfer, prepare: @@ -193,6 +197,7 @@ In [**Logs → Calls**](/observability/logs/call-logs), inspect the original cal | Symptom | Likely cause | Resolution | | --- | --- | --- | +| The transfer never starts an operator conversation, even though `warm-transfer-experimental` is configured | The assistant uses GPT-Live, which does not support assistant-based warm transfer | Use a standard assistant model, or configure the destination for a GPT-Live-compatible blind transfer on Twilio or Vapi SIP. | | The customer is transferred without an operator conversation | The destination uses a blind or non-assistant warm-transfer mode | Set `transferPlan.mode` to `warm-transfer-experimental`. | | The transfer assistant never speaks | `firstMessage` is omitted or `firstMessageMode` waits for the operator | Set `firstMessage` and use `assistant-speaks-first`, or let the operator speak first. | | The transfer assistant talks but never connects the parties | The prompt does not require `transferSuccessful` | Tell the transfer assistant to call `transferSuccessful` immediately after the operator accepts. | From 7da9a36cdf4637df10716bcbd8e2fca143b863c7 Mon Sep 17 00:00:00 2001 From: Stephen Smith Date: Mon, 19 Oct 2026 11:14:52 -0700 Subject: [PATCH 2/2] Remove GPT-Live troubleshooting row Co-Authored-By: Claude Opus 5.5 --- fern/calls/assistant-based-warm-transfer.mdx | 1 - 1 file changed, 1 deletion(-) diff --git a/fern/calls/assistant-based-warm-transfer.mdx b/fern/calls/assistant-based-warm-transfer.mdx index 98cdfd5b1..ad891fa1e 100644 --- a/fern/calls/assistant-based-warm-transfer.mdx +++ b/fern/calls/assistant-based-warm-transfer.mdx @@ -197,7 +197,6 @@ In [**Logs → Calls**](/observability/logs/call-logs), inspect the original cal | Symptom | Likely cause | Resolution | | --- | --- | --- | -| The transfer never starts an operator conversation, even though `warm-transfer-experimental` is configured | The assistant uses GPT-Live, which does not support assistant-based warm transfer | Use a standard assistant model, or configure the destination for a GPT-Live-compatible blind transfer on Twilio or Vapi SIP. | | The customer is transferred without an operator conversation | The destination uses a blind or non-assistant warm-transfer mode | Set `transferPlan.mode` to `warm-transfer-experimental`. | | The transfer assistant never speaks | `firstMessage` is omitted or `firstMessageMode` waits for the operator | Set `firstMessage` and use `assistant-speaks-first`, or let the operator speak first. | | The transfer assistant talks but never connects the parties | The prompt does not require `transferSuccessful` | Tell the transfer assistant to call `transferSuccessful` immediately after the operator accepts. |