@picsart/ai-sdk 5.9.0 → 5.10.1

package/_vendor/workflows-types/index.d.ts

@@ -6522,10 +6522,11 @@ interface XAiTtsResult {
 
 interface XAiImagesGenerationsCommand {
     prompt: string;
-    model?: "grok-imagine-image" | "grok-imagine-image-quality";
+    model?: "grok-imagine-image" | "grok-imagine-image-quality" | "grok-imagine-image-2.0";
     aspect_ratio?: "1:1" | "3:4" | "4:3" | "9:16" | "16:9" | "2:3" | "3:2" | "9:19.5" | "19.5:9" | "9:20" | "20:9" | "1:2" | "2:1" | "auto";
     n?: number;
     resolution?: "1k" | "2k";
+    quality?: "low" | "medium";
     options?: GenAIOptions$1a;
 }
 interface GenAIOptions$1a {
@@ -6564,7 +6565,7 @@ interface XAiImagesEditsCommand {
     image?: XAiImageUrl$1;
     images?: XAiImageUrl$1[];
     mask?: XAiImageUrl$1;
-    model?: "grok-imagine-image" | "grok-imagine-image-quality";
+    model?: "grok-imagine-image" | "grok-imagine-image-quality" | "grok-imagine-image-2.0";
     n?: number;
     resolution?: "1k" | "2k";
     options?: GenAIOptions$19;
@@ -9365,8 +9366,12 @@ interface AsyncCatalogVoice {
     name: string;
     description?: string;
     tags: string[];
+    preview?: AsyncCatalogVoicePreview;
     meta?: Record<string, unknown>;
 }
+interface AsyncCatalogVoicePreview {
+    audioUrl?: string;
+}
 
 interface GeminiCatalogVoicesCommand {
     cursor?: string;

package/index.d.ts

@@ -251,6 +251,11 @@ type ModelInputById = {
         imageUrls?: string[];
         thinking?: "off" | "low" | "medium" | "high";
     };
+    "gemini-3.7-flash": {
+        prompt: string;
+        imageUrls?: string[];
+        thinking?: "off" | "low" | "medium" | "high";
+    };
     "gemini-omni-flash-preview": {
         prompt: string;
         aspectRatio?: "16:9" | "9:16";
@@ -1523,7 +1528,7 @@ type ModelInputById = {
 type TypedModelId = keyof ModelInputById;
 type ModelInput<M extends TypedModelId> = ModelInputById[M];
 /** IDs of text-generation (LLM) models — narrows generateText(). */
-type TextModelId = "claude-haiku-4-5" | "claude-opus-4-8" | "claude-sonnet-4-6" | "gemini-3-pro" | "gemini-3.5-flash-lite" | "gemini-3.6-flash" | "gpt-5.5";
+type TextModelId = "claude-haiku-4-5" | "claude-opus-4-8" | "claude-sonnet-4-6" | "gemini-3-pro" | "gemini-3.5-flash-lite" | "gemini-3.6-flash" | "gemini-3.7-flash" | "gpt-5.5";
 type TextModelInputById = Pick<ModelInputById, TextModelId>;
 
 /**
@@ -2466,6 +2471,7 @@ declare const Models: {
     readonly Gemini31FlashLiteImage: "gemini-3.1-flash-lite-image";
     readonly Gemini35FlashLite: "gemini-3.5-flash-lite";
     readonly Gemini36Flash: "gemini-3.6-flash";
+    readonly Gemini37Flash: "gemini-3.7-flash";
     readonly GeminiOmniFlashPreview: "gemini-omni-flash-preview";
     readonly Gpt55: "gpt-5.5";
     readonly GptImage1: "gpt-image-1";

package/index.js

@@ -7772,7 +7772,7 @@ var { MODELS: MODELS32 } = defineModels("google", [
     description: "Extended music generation up to 184s with vocals, powered by Google Lyria 3 Pro.",
     features: [feat("Image Input", "input"), feat("Vocal & Instrumental", "characteristic"), feat("Up to 184s", "duration")],
     paramConfig: {
-      ...params.prompt(),
+      ...params.prompt({ placeholder: "Generate voiceover, music and sound effects" }),
       ...params.imageInput(10, "Mood Images")
     }
   }
@@ -8264,6 +8264,23 @@ var { MODELS: GEMINI_LLM } = defineModels("google", [
       ...thinkingParam(["low", "high"])
     }
   },
+  {
+    id: "gemini-3.7-flash",
+    name: "Gemini 3.7 Flash",
+    workflow: "chat-completions",
+    addedAt: "2026-08-19",
+    estimatedTime: 5,
+    mode: "text",
+    inputType: "i2t",
+    badge: ["fast"],
+    description: "Latest fast Gemini model \u2014 low-latency multimodal text generation.",
+    features: [feat("Vision", "input"), feat("Thinking", "characteristic")],
+    paramConfig: {
+      ...params.prompt(),
+      ...params.imageInput(8, "Images"),
+      ...thinkingParam(["low", "medium", "high"])
+    }
+  },
   {
     id: "gemini-3.6-flash",
     name: "Gemini 3.6 Flash",
@@ -8357,6 +8374,7 @@ registerPayloads(MODELS36, {
   "gemini-3-pro": buildGeminiPayload("gemini-3-pro-preview"),
   // Flash models route through chat-completions (OpenAI-shaped), not the
   // native `gemini` workflow. flash-lite has no thinking param → reasoning_effort omitted.
+  "gemini-3.7-flash": buildOpenAiPayload("gemini-3.7-flash"),
   "gemini-3.6-flash": buildOpenAiPayload("gemini-3.6-flash"),
   "gemini-3.5-flash-lite": buildOpenAiPayload("gemini-3.5-flash-lite")
 });
@@ -10496,6 +10514,7 @@ var Gemini31FlashImage = "gemini-3.1-flash-image";
 var Gemini31FlashLiteImage = "gemini-3.1-flash-lite-image";
 var Gemini35FlashLite = "gemini-3.5-flash-lite";
 var Gemini36Flash = "gemini-3.6-flash";
+var Gemini37Flash = "gemini-3.7-flash";
 var GeminiOmniFlashPreview = "gemini-omni-flash-preview";
 var Gpt55 = "gpt-5.5";
 var GptImage1 = "gpt-image-1";
@@ -10705,6 +10724,7 @@ var Models = {
   Gemini31FlashLiteImage,
   Gemini35FlashLite,
   Gemini36Flash,
+  Gemini37Flash,
   GeminiOmniFlashPreview,
   Gpt55,
   GptImage1,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@picsart/ai-sdk",
-  "version": "5.9.0",
+  "version": "5.10.1",
   "type": "module",
   "description": "Type-safe SDK for 100+ AI models — image, video, audio, and text generation with Picsart",
   "license": "MIT",