Align Vidu defaults and S2V image handling with tested server behavior

The server now keeps Vidu URL-based image2video requests on a valid default duration when callers omit duration, and it forwards MiniMax S2V reference images in the images[] shape expected by the deployed Hailuo adapter.

Constraint: The test environment currently validates provider behavior through popiartServer binary deploys rather than waiting for broader upstream adapter rollouts
Rejected: Keep the old shared 4-second default for all providers | breaks viduq3-turbo when duration is omitted
Rejected: Translate S2V references into subject_reference in popiartServer | the deployed Hailuo adapter already expects top-level images and performs its own mapping
Confidence: high
Scope-risk: narrow
Reversibility: clean
Directive: Keep provider-specific video defaults and reference-shape compatibility isolated in popiartServer unless upstream adapters are upgraded in lockstep
Tested: go test ./internal/server -run TestInferRouteKeyForModelRecognizesViduAsVideo|TestGenerate|TestResolve|TestSubmitMiniMaxVideoTask|TestSubmitImageToVideoTask|TestExecuteImageToImageJobUsesGenerationsPathForMiniMax|TestExecuteImageToVideoJob|TestResolveVideoReferencesSupportsImagesArray
Not-tested: Full PopiNewAPI unit suite from the local checkout

internal/server/newapi.go

@@ -825,10 +825,12 @@ func (c *newAPIClient) submitImageToVideoTask(ctx context.Context, token, modelI
 	if err := writer.WriteField("prompt", prompt); err != nil {
 		return "", err
 	}
-	if seconds := resolveVideoDurationSeconds(input); seconds != "" {
+	seconds := strings.TrimSpace(resolveVideoDurationSeconds(input))
-		if err := writer.WriteField("seconds", seconds); err != nil {
+	if seconds == "" {
-			return "", err
+		seconds = "4"
-		}
+	}
+	if err := writer.WriteField("seconds", seconds); err != nil {
+		return "", err
 	}
 	if size := resolveVideoSize(modelID, input, ref); size != "" {
 		if err := writer.WriteField("size", size); err != nil {
@@ -1435,7 +1437,7 @@ func filenameFromURL(rawURL, fallback string) string {
 
 func resolveVideoDurationSeconds(input map[string]any) string {
 	if input == nil {
-		return "4"
+		return ""
 	}
 	for _, value := range []any{input["seconds"], input["duration_s"], input["duration"]} {
 		switch typed := value.(type) {
@@ -1462,7 +1464,7 @@ func resolveVideoDurationSeconds(input map[string]any) string {
 			}
 		}
 	}
-	return "4"
+	return ""
 }
 
 func resolveVideoSize(modelID string, input map[string]any, ref imageEditReference) string {

internal/server/video_test.go

@@ -138,6 +138,35 @@ func TestSubmitImageToVideoTaskUsesURLImagesForViduModels(t *testing.T) {
 	}
 }
 
+func TestSubmitImageToVideoTaskUsesViduDefaultDurationWhenUnset(t *testing.T) {
+	var gotBody map[string]any
+
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		if err := json.NewDecoder(r.Body).Decode(&gotBody); err != nil {
+			t.Fatalf("decode request body: %v", err)
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"id":"task_vidu_default_duration","status":"queued"}`))
+	}))
+	defer srv.Close()
+
+	client := newNewAPIClient(Config{NewAPIBaseURL: srv.URL})
+	taskID, err := client.submitImageToVideoTask(context.Background(), "sk-test", "viduq3-turbo", map[string]any{
+		"prompt": "animate this still gently",
+	}, imageEditReference{
+		URL: "https://media.popi.test/m/demo/reference.png",
+	})
+	if err != nil {
+		t.Fatalf("submitImageToVideoTask: %v", err)
+	}
+	if taskID != "task_vidu_default_duration" {
+		t.Fatalf("unexpected task id: %q", taskID)
+	}
+	if gotBody["duration"] != float64(5) {
+		t.Fatalf("expected default vidu duration 5, got %#v", gotBody["duration"])
+	}
+}
+
 func TestSubmitMiniMaxVideoTaskUsesVideoGenerationsEndpoint(t *testing.T) {
 	var gotPath string
 	var gotBody map[string]any