diff --git a/common/model.go b/common/model.go
index 4ebc7b532d74..0f5957f071b8 100644
--- a/common/model.go
+++ b/common/model.go
@@ -12,7 +12,8 @@ var (
 	ImageGenerationModels = []string{
 		"dall-e-3",
 		"dall-e-2",
-		"gpt-image-1",
+		"prefix:gpt-image-",
+		"chatgpt-image-latest",
 		"prefix:imagen-",
 		"flux-",
 		"flux.1-",
diff --git a/relay/helper/text_endpoint_compatibility_test.go b/relay/helper/text_endpoint_compatibility_test.go
new file mode 100644
index 000000000000..69bc0aa6cd0e
--- /dev/null
+++ b/relay/helper/text_endpoint_compatibility_test.go
@@ -0,0 +1,62 @@
+package helper
+
+import (
+	"bytes"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	relayconstant "github.com/QuantumNous/new-api/relay/constant"
+	"github.com/gin-gonic/gin"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestChatCompletionsRejectsImageGenerationModels(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	tests := []struct {
+		name  string
+		model string
+	}{
+		{name: "current GPT Image model", model: "gpt-image-2"},
+		{name: "GPT Image snapshot", model: "gpt-image-2-2026-04-21"},
+		{name: "future GPT Image model", model: "gpt-image-3"},
+		{name: "legacy GPT Image model", model: "gpt-image-1.5"},
+		{name: "ChatGPT image model", model: "chatgpt-image-latest"},
+		{name: "DALL-E model", model: "dall-e-3"},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			body := `{"model":"` + tt.model + `","messages":[{"role":"user","content":"Generate an image"}]}`
+			c, _ := gin.CreateTestContext(httptest.NewRecorder())
+			c.Request = httptest.NewRequest(http.MethodPost, "/v1/chat/completions", bytes.NewBufferString(body))
+			c.Request.Header.Set("Content-Type", "application/json")
+
+			_, err := GetAndValidateTextRequest(c, relayconstant.RelayModeChatCompletions)
+
+			require.Error(t, err)
+			assert.Equal(
+				t,
+				`model "`+tt.model+`" is not supported on /v1/chat/completions; use /v1/images/generations instead`,
+				err.Error(),
+			)
+		})
+	}
+}
+
+func TestChatCompletionsAcceptsTextModel(t *testing.T) {
+	c, _ := gin.CreateTestContext(httptest.NewRecorder())
+	c.Request = httptest.NewRequest(
+		http.MethodPost,
+		"/v1/chat/completions",
+		bytes.NewBufferString(`{"model":"gpt-5","messages":[{"role":"user","content":"Hello"}]}`),
+	)
+	c.Request.Header.Set("Content-Type", "application/json")
+
+	request, err := GetAndValidateTextRequest(c, relayconstant.RelayModeChatCompletions)
+
+	require.NoError(t, err)
+	assert.Equal(t, "gpt-5", request.Model)
+}
diff --git a/relay/helper/valid_request.go b/relay/helper/valid_request.go
index 22068a45ca5c..4b5859bdb621 100644
--- a/relay/helper/valid_request.go
+++ b/relay/helper/valid_request.go
@@ -329,6 +329,12 @@ func GetAndValidateTextRequest(c *gin.Context, relayMode int) (*dto.GeneralOpenA
 	if textRequest.Model == "" {
 		return nil, errors.New("model is required")
 	}
+	if relayMode == relayconstant.RelayModeChatCompletions && common.IsImageGenerationModel(textRequest.Model) {
+		return nil, fmt.Errorf(
+			"model %q is not supported on /v1/chat/completions; use /v1/images/generations instead",
+			textRequest.Model,
+		)
+	}
 	if textRequest.WebSearchOptions != nil {
 		if textRequest.WebSearchOptions.SearchContextSize != "" {
 			validSizes := map[string]bool{