diff --git a/pkg/lmcli/provider/anthropic/anthropic.go b/pkg/lmcli/provider/anthropic/anthropic.go
index 706f449..51586c4 100644
--- a/pkg/lmcli/provider/anthropic/anthropic.go
+++ b/pkg/lmcli/provider/anthropic/anthropic.go
@@ -229,8 +229,7 @@ func (c *AnthropicClient) CreateChatCompletionStream(
 			case "message_start":
 				// noop
 			case "ping":
-				// write an empty string to signal start of text
-				output <- ""
+				// signals start of text - currently ignoring
 			case "content_block_start":
 				// ignore?
 			case "content_block_delta":
diff --git a/pkg/tui/views/chat/chat.go b/pkg/tui/views/chat/chat.go
index 4295476..cd69232 100644
--- a/pkg/tui/views/chat/chat.go
+++ b/pkg/tui/views/chat/chat.go
@@ -278,6 +278,8 @@ func (m Model) Update(msg tea.Msg) (Model, tea.Cmd) {
 		m.updateContent()
 		m.content.GotoBottom()
 	case msgResponseChunk:
+		cmds = append(cmds, m.waitForChunk()) // wait for the next chunk
+
 		chunk := string(msg)
 		if chunk == "" {
 			break
@@ -295,7 +297,6 @@ func (m Model) Update(msg tea.Msg) (Model, tea.Cmd) {
 			})
 		}
 		m.updateContent()
-		cmds = append(cmds, m.waitForChunk()) // wait for the next chunk
 
 		// show cursor and reset blink interval (simulate typing)
 		m.replyCursor.Blink = false
@@ -304,7 +305,8 @@ func (m Model) Update(msg tea.Msg) (Model, tea.Cmd) {
 		m.tokenCount++
 		m.elapsed = time.Now().Sub(m.startTime)
 	case msgAssistantReply:
-		// the last reply that was being worked on is finished
+		cmds = append(cmds, m.waitForReply()) // wait for the next reply
+
 		reply := models.Message(msg)
 		reply.Content = strings.TrimSpace(reply.Content)
 
@@ -332,7 +334,6 @@ func (m Model) Update(msg tea.Msg) (Model, tea.Cmd) {
 		}
 
 		m.updateContent()
-		cmds = append(cmds, m.waitForReply())
 	case msgResponseEnd:
 		m.waitingForReply = false
 		last := len(m.messages) - 1