@huggingface/inference 4.13.4 → 4.13.6

package/dist/esm/providers/zai-org.d.ts

@@ -1,24 +1,22 @@
-/**
- * See the registered mapping of HF model ID => ZAI model ID here:
- *
- * https://huggingface.co/api/partners/zai-org/models
- *
- * This is a publicly available mapping.
- *
- * If you want to try to run inference for a new model locally before it's registered on huggingface.co,
- * you can add it to the dictionary "HARDCODED_MODEL_ID_MAPPING" in consts.ts, for dev purposes.
- *
- * - If you work at zai and want to update this mapping, please use the model mapping API we provide on huggingface.co
- * - If you're a community member and want to add a new supported HF model to zai, please open an issue on the present repo
- * and we will tag zai team members.
- *
- * Thanks!
- */
-import { BaseConversationalTask } from "./providerHelper.js";
-import type { HeaderParams } from "../types.js";
+import type { BodyParams, HeaderParams } from "../types.js";
+import { BaseConversationalTask, TaskProviderHelper, type TextToImageTaskHelper } from "./providerHelper.js";
 export declare class ZaiConversationalTask extends BaseConversationalTask {
     constructor();
     prepareHeaders(params: HeaderParams, binary: boolean): Record<string, string>;
     makeRoute(): string;
 }
+interface ZaiTextToImageResponse {
+    model: string;
+    id: string;
+    request_id: string;
+    task_status: "PROCESSING" | "SUCCESS" | "FAIL";
+}
+export declare class ZaiTextToImageTask extends TaskProviderHelper implements TextToImageTaskHelper {
+    constructor();
+    prepareHeaders(params: HeaderParams, binary: boolean): Record<string, string>;
+    makeRoute(): string;
+    preparePayload(params: BodyParams): Record<string, unknown>;
+    getResponse(response: ZaiTextToImageResponse, url?: string, headers?: Record<string, string>, outputType?: "url" | "blob" | "json"): Promise<string | Blob | Record<string, unknown>>;
+}
+export {};
 //# sourceMappingURL=zai-org.d.ts.map

package/dist/esm/providers/zai-org.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"zai-org.d.ts","sourceRoot":"","sources":["../../../src/providers/zai-org.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,OAAO,EAAE,sBAAsB,EAAE,MAAM,qBAAqB,CAAC;AAC7D,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAIhD,qBAAa,qBAAsB,SAAQ,sBAAsB;;IAKvD,cAAc,CAAC,MAAM,EAAE,YAAY,EAAE,MAAM,EAAE,OAAO,GAAG,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAO7E,SAAS,IAAI,MAAM;CAG5B"}
+{"version":3,"file":"zai-org.d.ts","sourceRoot":"","sources":["../../../src/providers/zai-org.ts"],"names":[],"mappings":"AAkBA,OAAO,KAAK,EAAE,UAAU,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAG5D,OAAO,EAAE,sBAAsB,EAAE,kBAAkB,EAAE,KAAK,qBAAqB,EAAE,MAAM,qBAAqB,CAAC;AAI7G,qBAAa,qBAAsB,SAAQ,sBAAsB;;IAKvD,cAAc,CAAC,MAAM,EAAE,YAAY,EAAE,MAAM,EAAE,OAAO,GAAG,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAO7E,SAAS,IAAI,MAAM;CAG5B;AAED,UAAU,sBAAsB;IAC/B,KAAK,EAAE,MAAM,CAAC;IACd,EAAE,EAAE,MAAM,CAAC;IACX,UAAU,EAAE,MAAM,CAAC;IACnB,WAAW,EAAE,YAAY,GAAG,SAAS,GAAG,MAAM,CAAC;CAC/C;AAaD,qBAAa,kBAAmB,SAAQ,kBAAmB,YAAW,qBAAqB;;IAKjF,cAAc,CAAC,MAAM,EAAE,YAAY,EAAE,MAAM,EAAE,OAAO,GAAG,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAYtF,SAAS,IAAI,MAAM;IAInB,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;IASrD,WAAW,CAChB,QAAQ,EAAE,sBAAsB,EAChC,GAAG,CAAC,EAAE,MAAM,EACZ,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,EAChC,UAAU,CAAC,EAAE,KAAK,GAAG,MAAM,GAAG,MAAM,GAClC,OAAO,CAAC,MAAM,GAAG,IAAI,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CAsFnD"}

package/dist/esm/providers/zai-org.js

@@ -14,7 +14,11 @@
  *
  * Thanks!
  */
-import { BaseConversationalTask } from "./providerHelper.js";
+import { InferenceClientProviderApiError, InferenceClientProviderOutputError } from "../errors.js";
+import { isUrl } from "../lib/isUrl.js";
+import { delay } from "../utils/delay.js";
+import { omit } from "../utils/omit.js";
+import { BaseConversationalTask, TaskProviderHelper } from "./providerHelper.js";
 const ZAI_API_BASE_URL = "https://api.z.ai";
 export class ZaiConversationalTask extends BaseConversationalTask {
     constructor() {
@@ -30,3 +34,87 @@ export class ZaiConversationalTask extends BaseConversationalTask {
         return "/api/paas/v4/chat/completions";
     }
 }
+const MAX_POLL_ATTEMPTS = 60;
+const POLL_INTERVAL_MS = 5000;
+export class ZaiTextToImageTask extends TaskProviderHelper {
+    constructor() {
+        super("zai-org", ZAI_API_BASE_URL);
+    }
+    prepareHeaders(params, binary) {
+        const headers = {
+            Authorization: `Bearer ${params.accessToken}`,
+            "x-source-channel": "hugging_face",
+            "accept-language": "en-US,en",
+        };
+        if (!binary) {
+            headers["Content-Type"] = "application/json";
+        }
+        return headers;
+    }
+    makeRoute() {
+        return "/api/paas/v4/async/images/generations";
+    }
+    preparePayload(params) {
+        return {
+            ...omit(params.args, ["inputs", "parameters"]),
+            ...params.args.parameters,
+            model: params.model,
+            prompt: params.args.inputs,
+        };
+    }
+    async getResponse(response, url, headers, outputType) {
+        void url;
+        if (typeof response !== "object" ||
+            !response ||
+            !("task_status" in response) ||
+            !("id" in response) ||
+            typeof response.id !== "string") {
+            throw new InferenceClientProviderOutputError(`Received malformed response from ZAI text-to-image API: expected { id: string, task_status: string }, got: ${JSON.stringify(response)}`);
+        }
+        if (response.task_status === "FAIL") {
+            throw new InferenceClientProviderOutputError("ZAI API returned task status: FAIL");
+        }
+        const taskId = response.id;
+        const pollUrl = `${ZAI_API_BASE_URL}/api/paas/v4/async-result/${taskId}`;
+        const pollHeaders = {
+            "x-source-channel": "hugging_face",
+            "accept-language": "en-US,en",
+        };
+        if (headers && headers["Authorization"]) {
+            pollHeaders["Authorization"] = headers["Authorization"];
+        }
+        for (let attempt = 0; attempt < MAX_POLL_ATTEMPTS; attempt++) {
+            await delay(POLL_INTERVAL_MS);
+            const resp = await fetch(pollUrl, {
+                method: "GET",
+                headers: pollHeaders,
+            });
+            if (!resp.ok) {
+                throw new InferenceClientProviderApiError(`Failed to fetch result from ZAI text-to-image API: ${resp.status}`, { url: pollUrl, method: "GET" }, { requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: await resp.text() });
+            }
+            const result = await resp.json();
+            if (result.task_status === "FAIL") {
+                throw new InferenceClientProviderOutputError("ZAI text-to-image API task failed");
+            }
+            if (result.task_status === "SUCCESS") {
+                if (!result.image_result ||
+                    !Array.isArray(result.image_result) ||
+                    result.image_result.length === 0 ||
+                    typeof result.image_result[0]?.url !== "string" ||
+                    !isUrl(result.image_result[0].url)) {
+                    throw new InferenceClientProviderOutputError(`Received malformed response from ZAI text-to-image API: expected { image_result: Array<{ url: string }> }, got: ${JSON.stringify(result)}`);
+                }
+                const imageUrl = result.image_result[0].url;
+                if (outputType === "json") {
+                    return { ...result };
+                }
+                if (outputType === "url") {
+                    return imageUrl;
+                }
+                const imageResponse = await fetch(imageUrl);
+                return await imageResponse.blob();
+            }
+        }
+        throw new InferenceClientProviderOutputError(`Timed out while waiting for the result from ZAI API - aborting after ${MAX_POLL_ATTEMPTS} attempts`);
+    }
+}

package/dist/esm/snippets/getInferenceSnippets.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"getInferenceSnippets.d.ts","sourceRoot":"","sources":["../../../src/snippets/getInferenceSnippets.ts"],"names":[],"mappings":"AACA,OAAO,EACN,KAAK,gBAAgB,EAErB,KAAK,gBAAgB,EAGrB,MAAM,oBAAoB,CAAC;AAK5B,OAAO,KAAK,EAAE,6BAA6B,EAAE,yBAAyB,EAA8B,MAAM,aAAa,CAAC;AAKxH,MAAM,MAAM,uBAAuB,GAAG;IACrC,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,MAAM,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACjC,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;AAuY5B,wBAAgB,oBAAoB,CACnC,KAAK,EAAE,gBAAgB,EACvB,QAAQ,EAAE,yBAAyB,EACnC,wBAAwB,CAAC,EAAE,6BAA6B,EACxD,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,EAAE,CAIpB"}
+{"version":3,"file":"getInferenceSnippets.d.ts","sourceRoot":"","sources":["../../../src/snippets/getInferenceSnippets.ts"],"names":[],"mappings":"AACA,OAAO,EACN,KAAK,gBAAgB,EAErB,KAAK,gBAAgB,EAGrB,MAAM,oBAAoB,CAAC;AAK5B,OAAO,KAAK,EAAE,6BAA6B,EAAE,yBAAyB,EAA8B,MAAM,aAAa,CAAC;AAKxH,MAAM,MAAM,uBAAuB,GAAG;IACrC,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,MAAM,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACjC,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;AA4Y5B,wBAAgB,oBAAoB,CACnC,KAAK,EAAE,gBAAgB,EACvB,QAAQ,EAAE,yBAAyB,EACnC,wBAAwB,CAAC,EAAE,6BAA6B,EACxD,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,EAAE,CAIpB"}

package/dist/esm/snippets/getInferenceSnippets.js

@@ -41,7 +41,10 @@ const HF_PYTHON_METHODS = {
     "image-classification": "image_classification",
     "image-segmentation": "image_segmentation",
     "image-to-image": "image_to_image",
+    "image-to-video": "image_to_video",
     "image-to-text": "image_to_text",
+    "image-text-to-image": "image_text_to_image",
+    "image-text-to-video": "image_text_to_video",
     "object-detection": "object_detection",
     "question-answering": "question_answering",
     "sentence-similarity": "sentence_similarity",
@@ -274,7 +277,9 @@ const snippets = {
     "fill-mask": snippetGenerator("basic"),
     "image-classification": snippetGenerator("basicImage"),
     "image-segmentation": snippetGenerator("basicImage"),
+    "image-text-to-image": snippetGenerator("imageToImage", prepareImageToImageInput),
     "image-text-to-text": snippetGenerator("conversational"),
+    "image-text-to-video": snippetGenerator("imageToVideo", prepareImageToImageInput),
     "image-to-image": snippetGenerator("imageToImage", prepareImageToImageInput),
     "image-to-text": snippetGenerator("basicImage"),
     "image-to-video": snippetGenerator("imageToVideo", prepareImageToImageInput),

package/dist/esm/tasks/cv/imageTextToImage.d.ts

@@ -0,0 +1,9 @@
+import type { ImageTextToImageInput } from "@huggingface/tasks";
+import type { BaseArgs, Options } from "../../types.js";
+export type ImageTextToImageArgs = BaseArgs & ImageTextToImageInput;
+/**
+ * This task takes an image and text input and outputs a new generated image.
+ * Recommended model: black-forest-labs/FLUX.2-dev
+ */
+export declare function imageTextToImage(args: ImageTextToImageArgs, options?: Options): Promise<Blob>;
+//# sourceMappingURL=imageTextToImage.d.ts.map

package/dist/esm/tasks/cv/imageTextToImage.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"imageTextToImage.d.ts","sourceRoot":"","sources":["../../../../src/tasks/cv/imageTextToImage.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,qBAAqB,EAAE,MAAM,oBAAoB,CAAC;AAGhE,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,MAAM,gBAAgB,CAAC;AAGxD,MAAM,MAAM,oBAAoB,GAAG,QAAQ,GAAG,qBAAqB,CAAC;AAEpE;;;GAGG;AACH,wBAAsB,gBAAgB,CAAC,IAAI,EAAE,oBAAoB,EAAE,OAAO,CAAC,EAAE,OAAO,GAAG,OAAO,CAAC,IAAI,CAAC,CASnG"}

package/dist/esm/tasks/cv/imageTextToImage.js

@@ -0,0 +1,17 @@
+import { resolveProvider } from "../../lib/getInferenceProviderMapping.js";
+import { getProviderHelper } from "../../lib/getProviderHelper.js";
+import { innerRequest } from "../../utils/request.js";
+/**
+ * This task takes an image and text input and outputs a new generated image.
+ * Recommended model: black-forest-labs/FLUX.2-dev
+ */
+export async function imageTextToImage(args, options) {
+    const provider = await resolveProvider(args.provider, args.model, args.endpointUrl);
+    const providerHelper = getProviderHelper(provider, "image-text-to-image");
+    const payload = await providerHelper.preparePayloadAsync(args);
+    const { data: res, requestContext } = await innerRequest(payload, providerHelper, {
+        ...options,
+        task: "image-text-to-image",
+    });
+    return providerHelper.getResponse(res, requestContext.url, requestContext.info.headers);
+}

package/dist/esm/tasks/cv/imageTextToVideo.d.ts

@@ -0,0 +1,9 @@
+import type { ImageTextToVideoInput } from "@huggingface/tasks";
+import type { BaseArgs, Options } from "../../types.js";
+export type ImageTextToVideoArgs = BaseArgs & ImageTextToVideoInput;
+/**
+ * This task takes an image and text input and outputs a generated video.
+ * Recommended model: Lightricks/LTX-Video
+ */
+export declare function imageTextToVideo(args: ImageTextToVideoArgs, options?: Options): Promise<Blob>;
+//# sourceMappingURL=imageTextToVideo.d.ts.map

package/dist/esm/tasks/cv/imageTextToVideo.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"imageTextToVideo.d.ts","sourceRoot":"","sources":["../../../../src/tasks/cv/imageTextToVideo.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,qBAAqB,EAAE,MAAM,oBAAoB,CAAC;AAGhE,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,MAAM,gBAAgB,CAAC;AAGxD,MAAM,MAAM,oBAAoB,GAAG,QAAQ,GAAG,qBAAqB,CAAC;AAEpE;;;GAGG;AACH,wBAAsB,gBAAgB,CAAC,IAAI,EAAE,oBAAoB,EAAE,OAAO,CAAC,EAAE,OAAO,GAAG,OAAO,CAAC,IAAI,CAAC,CASnG"}

package/dist/esm/tasks/cv/imageTextToVideo.js

@@ -0,0 +1,17 @@
+import { resolveProvider } from "../../lib/getInferenceProviderMapping.js";
+import { getProviderHelper } from "../../lib/getProviderHelper.js";
+import { innerRequest } from "../../utils/request.js";
+/**
+ * This task takes an image and text input and outputs a generated video.
+ * Recommended model: Lightricks/LTX-Video
+ */
+export async function imageTextToVideo(args, options) {
+    const provider = await resolveProvider(args.provider, args.model, args.endpointUrl);
+    const providerHelper = getProviderHelper(provider, "image-text-to-video");
+    const payload = await providerHelper.preparePayloadAsync(args);
+    const { data: res, requestContext } = await innerRequest(payload, providerHelper, {
+        ...options,
+        task: "image-text-to-video",
+    });
+    return providerHelper.getResponse(res, requestContext.url, requestContext.info.headers);
+}

package/dist/esm/tasks/index.d.ts

@@ -9,6 +9,8 @@ export * from "./cv/imageSegmentation.js";
 export * from "./cv/imageToImage.js";
 export * from "./cv/imageToText.js";
 export * from "./cv/imageToVideo.js";
+export * from "./cv/imageTextToImage.js";
+export * from "./cv/imageTextToVideo.js";
 export * from "./cv/objectDetection.js";
 export * from "./cv/textToImage.js";
 export * from "./cv/textToVideo.js";

package/dist/esm/tasks/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AACA,cAAc,qBAAqB,CAAC;AACpC,cAAc,8BAA8B,CAAC;AAG7C,cAAc,gCAAgC,CAAC;AAC/C,cAAc,yBAAyB,CAAC;AACxC,cAAc,uCAAuC,CAAC;AACtD,cAAc,yBAAyB,CAAC;AAGxC,cAAc,6BAA6B,CAAC;AAC5C,cAAc,2BAA2B,CAAC;AAC1C,cAAc,sBAAsB,CAAC;AACrC,cAAc,qBAAqB,CAAC;AACpC,cAAc,sBAAsB,CAAC;AACrC,cAAc,yBAAyB,CAAC;AACxC,cAAc,qBAAqB,CAAC;AACpC,cAAc,qBAAqB,CAAC;AACpC,cAAc,qCAAqC,CAAC;AAGpD,cAAc,yBAAyB,CAAC;AACxC,cAAc,+BAA+B,CAAC;AAC9C,cAAc,4BAA4B,CAAC;AAC3C,cAAc,mBAAmB,CAAC;AAClC,cAAc,4BAA4B,CAAC;AAC3C,cAAc,6BAA6B,CAAC;AAC5C,cAAc,wBAAwB,CAAC;AACvC,cAAc,iCAAiC,CAAC;AAChD,cAAc,6BAA6B,CAAC;AAC5C,cAAc,yBAAyB,CAAC;AACxC,cAAc,+BAA+B,CAAC;AAC9C,cAAc,8BAA8B,CAAC;AAC7C,cAAc,sBAAsB,CAAC;AACrC,cAAc,iCAAiC,CAAC;AAGhD,cAAc,2CAA2C,CAAC;AAC1D,cAAc,yCAAyC,CAAC;AAGxD,cAAc,oCAAoC,CAAC;AACnD,cAAc,gCAAgC,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AACA,cAAc,qBAAqB,CAAC;AACpC,cAAc,8BAA8B,CAAC;AAG7C,cAAc,gCAAgC,CAAC;AAC/C,cAAc,yBAAyB,CAAC;AACxC,cAAc,uCAAuC,CAAC;AACtD,cAAc,yBAAyB,CAAC;AAGxC,cAAc,6BAA6B,CAAC;AAC5C,cAAc,2BAA2B,CAAC;AAC1C,cAAc,sBAAsB,CAAC;AACrC,cAAc,qBAAqB,CAAC;AACpC,cAAc,sBAAsB,CAAC;AACrC,cAAc,0BAA0B,CAAC;AACzC,cAAc,0BAA0B,CAAC;AACzC,cAAc,yBAAyB,CAAC;AACxC,cAAc,qBAAqB,CAAC;AACpC,cAAc,qBAAqB,CAAC;AACpC,cAAc,qCAAqC,CAAC;AAGpD,cAAc,yBAAyB,CAAC;AACxC,cAAc,+BAA+B,CAAC;AAC9C,cAAc,4BAA4B,CAAC;AAC3C,cAAc,mBAAmB,CAAC;AAClC,cAAc,4BAA4B,CAAC;AAC3C,cAAc,6BAA6B,CAAC;AAC5C,cAAc,wBAAwB,CAAC;AACvC,cAAc,iCAAiC,CAAC;AAChD,cAAc,6BAA6B,CAAC;AAC5C,cAAc,yBAAyB,CAAC;AACxC,cAAc,+BAA+B,CAAC;AAC9C,cAAc,8BAA8B,CAAC;AAC7C,cAAc,sBAAsB,CAAC;AACrC,cAAc,iCAAiC,CAAC;AAGhD,cAAc,2CAA2C,CAAC;AAC1D,cAAc,yCAAyC,CAAC;AAGxD,cAAc,oCAAoC,CAAC;AACnD,cAAc,gCAAgC,CAAC"}

package/dist/esm/tasks/index.js

@@ -12,6 +12,8 @@ export * from "./cv/imageSegmentation.js";
 export * from "./cv/imageToImage.js";
 export * from "./cv/imageToText.js";
 export * from "./cv/imageToVideo.js";
+export * from "./cv/imageTextToImage.js";
+export * from "./cv/imageTextToVideo.js";
 export * from "./cv/objectDetection.js";
 export * from "./cv/textToImage.js";
 export * from "./cv/textToVideo.js";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@huggingface/inference",
-  "version": "4.13.4",
+  "version": "4.13.6",
   "license": "MIT",
   "author": "Hugging Face and Tim Mikeladze <tim.mikeladze@gmail.com>",
   "description": "Typescript client for the Hugging Face Inference Providers and Inference Endpoints",
@@ -40,8 +40,8 @@
   },
   "type": "module",
   "dependencies": {
-    "@huggingface/tasks": "^0.19.65",
-    "@huggingface/jinja": "^0.5.1"
+    "@huggingface/tasks": "^0.19.76",
+    "@huggingface/jinja": "^0.5.3"
   },
   "devDependencies": {
     "@types/node": "18.13.0"

package/src/lib/getProviderHelper.ts

@@ -28,6 +28,8 @@ import type {
 	ImageToImageTaskHelper,
 	ImageToTextTaskHelper,
 	ImageToVideoTaskHelper,
+	ImageTextToImageTaskHelper,
+	ImageTextToVideoTaskHelper,
 	ObjectDetectionTaskHelper,
 	QuestionAnsweringTaskHelper,
 	SentenceSimilarityTaskHelper,
@@ -182,6 +184,7 @@ export const PROVIDERS: Record<InferenceProvider, Partial<Record<InferenceTask,
 	},
 	"zai-org": {
 		conversational: new Zai.ZaiConversationalTask(),
+		"text-to-image": new Zai.ZaiTextToImageTask(),
 	},
 };
 
@@ -276,6 +279,14 @@ export function getProviderHelper(
 	provider: InferenceProviderOrPolicy,
 	task: "image-to-video"
 ): ImageToVideoTaskHelper & TaskProviderHelper;
+export function getProviderHelper(
+	provider: InferenceProviderOrPolicy,
+	task: "image-text-to-image"
+): ImageTextToImageTaskHelper & TaskProviderHelper;
+export function getProviderHelper(
+	provider: InferenceProviderOrPolicy,
+	task: "image-text-to-video"
+): ImageTextToVideoTaskHelper & TaskProviderHelper;
 export function getProviderHelper(
 	provider: InferenceProviderOrPolicy,
 	task: "sentence-similarity"

package/src/package.ts

@@ -1,3 +1,3 @@
 // Generated file from package.json. Issues importing JSON directly when publishing on commonjs/ESM - see https://github.com/microsoft/TypeScript/issues/51783
-export const PACKAGE_VERSION = "4.13.4";
+export const PACKAGE_VERSION = "4.13.6";
 export const PACKAGE_NAME = "@huggingface/inference";

package/src/providers/providerHelper.ts

@@ -19,6 +19,8 @@ import type {
 	ImageToTextInput,
 	ImageToTextOutput,
 	ImageToVideoInput,
+	ImageTextToImageInput,
+	ImageTextToVideoInput,
 	ObjectDetectionInput,
 	ObjectDetectionOutput,
 	QuestionAnsweringInput,
@@ -54,6 +56,8 @@ import { toArray } from "../utils/toArray.js";
 import type { ImageToImageArgs } from "../tasks/cv/imageToImage.js";
 import type { AutomaticSpeechRecognitionArgs } from "../tasks/audio/automaticSpeechRecognition.js";
 import type { ImageToVideoArgs } from "../tasks/cv/imageToVideo.js";
+import type { ImageTextToImageArgs } from "../tasks/cv/imageTextToImage.js";
+import type { ImageTextToVideoArgs } from "../tasks/cv/imageTextToVideo.js";
 import type { ImageSegmentationArgs } from "../tasks/cv/imageSegmentation.js";
 
 /**
@@ -159,6 +163,18 @@ export interface ImageToVideoTaskHelper {
 	preparePayloadAsync(args: ImageToVideoArgs): Promise<RequestArgs>;
 }
 
+export interface ImageTextToImageTaskHelper {
+	getResponse(response: unknown, url?: string, headers?: HeadersInit): Promise<Blob>;
+	preparePayload(params: BodyParams<ImageTextToImageInput & BaseArgs>): Record<string, unknown>;
+	preparePayloadAsync(args: ImageTextToImageArgs): Promise<RequestArgs>;
+}
+
+export interface ImageTextToVideoTaskHelper {
+	getResponse(response: unknown, url?: string, headers?: HeadersInit): Promise<Blob>;
+	preparePayload(params: BodyParams<ImageTextToVideoInput & BaseArgs>): Record<string, unknown>;
+	preparePayloadAsync(args: ImageTextToVideoArgs): Promise<RequestArgs>;
+}
+
 export interface ImageSegmentationTaskHelper {
 	getResponse(response: unknown, url?: string, headers?: HeadersInit): Promise<ImageSegmentationOutput>;
 	preparePayload(params: BodyParams<ImageSegmentationInput & BaseArgs>): Record<string, unknown> | BodyInit;

package/src/providers/zai-org.ts

@@ -14,8 +14,12 @@
  *
  * Thanks!
  */
-import { BaseConversationalTask } from "./providerHelper.js";
-import type { HeaderParams } from "../types.js";
+import { InferenceClientProviderApiError, InferenceClientProviderOutputError } from "../errors.js";
+import { isUrl } from "../lib/isUrl.js";
+import type { BodyParams, HeaderParams } from "../types.js";
+import { delay } from "../utils/delay.js";
+import { omit } from "../utils/omit.js";
+import { BaseConversationalTask, TaskProviderHelper, type TextToImageTaskHelper } from "./providerHelper.js";
 
 const ZAI_API_BASE_URL = "https://api.z.ai";
 
@@ -35,3 +39,144 @@ export class ZaiConversationalTask extends BaseConversationalTask {
 		return "/api/paas/v4/chat/completions";
 	}
 }
+
+interface ZaiTextToImageResponse {
+	model: string;
+	id: string;
+	request_id: string;
+	task_status: "PROCESSING" | "SUCCESS" | "FAIL";
+}
+
+interface ZaiAsyncResultResponse {
+	image_result?: Array<{ url: string }>;
+	model: string;
+	id: string;
+	request_id: string;
+	task_status: "PROCESSING" | "SUCCESS" | "FAIL";
+}
+
+const MAX_POLL_ATTEMPTS = 60;
+const POLL_INTERVAL_MS = 5000;
+
+export class ZaiTextToImageTask extends TaskProviderHelper implements TextToImageTaskHelper {
+	constructor() {
+		super("zai-org", ZAI_API_BASE_URL);
+	}
+
+	override prepareHeaders(params: HeaderParams, binary: boolean): Record<string, string> {
+		const headers: Record<string, string> = {
+			Authorization: `Bearer ${params.accessToken}`,
+			"x-source-channel": "hugging_face",
+			"accept-language": "en-US,en",
+		};
+		if (!binary) {
+			headers["Content-Type"] = "application/json";
+		}
+		return headers;
+	}
+
+	makeRoute(): string {
+		return "/api/paas/v4/async/images/generations";
+	}
+
+	preparePayload(params: BodyParams): Record<string, unknown> {
+		return {
+			...omit(params.args, ["inputs", "parameters"]),
+			...(params.args.parameters as Record<string, unknown>),
+			model: params.model,
+			prompt: params.args.inputs,
+		};
+	}
+
+	async getResponse(
+		response: ZaiTextToImageResponse,
+		url?: string,
+		headers?: Record<string, string>,
+		outputType?: "url" | "blob" | "json"
+	): Promise<string | Blob | Record<string, unknown>> {
+		void url;
+		if (
+			typeof response !== "object" ||
+			!response ||
+			!("task_status" in response) ||
+			!("id" in response) ||
+			typeof response.id !== "string"
+		) {
+			throw new InferenceClientProviderOutputError(
+				`Received malformed response from ZAI text-to-image API: expected { id: string, task_status: string }, got: ${JSON.stringify(
+					response
+				)}`
+			);
+		}
+
+		if (response.task_status === "FAIL") {
+			throw new InferenceClientProviderOutputError("ZAI API returned task status: FAIL");
+		}
+
+		const taskId = response.id;
+		const pollUrl = `${ZAI_API_BASE_URL}/api/paas/v4/async-result/${taskId}`;
+
+		const pollHeaders: Record<string, string> = {
+			"x-source-channel": "hugging_face",
+			"accept-language": "en-US,en",
+		};
+		if (headers && headers["Authorization"]) {
+			pollHeaders["Authorization"] = headers["Authorization"];
+		}
+
+		for (let attempt = 0; attempt < MAX_POLL_ATTEMPTS; attempt++) {
+			await delay(POLL_INTERVAL_MS);
+
+			const resp = await fetch(pollUrl, {
+				method: "GET",
+				headers: pollHeaders,
+			});
+
+			if (!resp.ok) {
+				throw new InferenceClientProviderApiError(
+					`Failed to fetch result from ZAI text-to-image API: ${resp.status}`,
+					{ url: pollUrl, method: "GET" },
+					{ requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: await resp.text() }
+				);
+			}
+
+			const result: ZaiAsyncResultResponse = await resp.json();
+
+			if (result.task_status === "FAIL") {
+				throw new InferenceClientProviderOutputError("ZAI text-to-image API task failed");
+			}
+
+			if (result.task_status === "SUCCESS") {
+				if (
+					!result.image_result ||
+					!Array.isArray(result.image_result) ||
+					result.image_result.length === 0 ||
+					typeof result.image_result[0]?.url !== "string" ||
+					!isUrl(result.image_result[0].url)
+				) {
+					throw new InferenceClientProviderOutputError(
+						`Received malformed response from ZAI text-to-image API: expected { image_result: Array<{ url: string }> }, got: ${JSON.stringify(
+							result
+						)}`
+					);
+				}
+
+				const imageUrl = result.image_result[0].url;
+
+				if (outputType === "json") {
+					return { ...result };
+				}
+				if (outputType === "url") {
+					return imageUrl;
+				}
+
+				const imageResponse = await fetch(imageUrl);
+				return await imageResponse.blob();
+			}
+		}
+
+		throw new InferenceClientProviderOutputError(
+			`Timed out while waiting for the result from ZAI API - aborting after ${MAX_POLL_ATTEMPTS} attempts`
+		);
+	}
+}

package/src/snippets/getInferenceSnippets.ts

@@ -96,7 +96,10 @@ const HF_PYTHON_METHODS: Partial<Record<WidgetType, string>> = {
 	"image-classification": "image_classification",
 	"image-segmentation": "image_segmentation",
 	"image-to-image": "image_to_image",
+	"image-to-video": "image_to_video",
 	"image-to-text": "image_to_text",
+	"image-text-to-image": "image_text_to_image",
+	"image-text-to-video": "image_text_to_video",
 	"object-detection": "object_detection",
 	"question-answering": "question_answering",
 	"sentence-similarity": "sentence_similarity",
@@ -390,7 +393,9 @@ const snippets: Partial<
 	"fill-mask": snippetGenerator("basic"),
 	"image-classification": snippetGenerator("basicImage"),
 	"image-segmentation": snippetGenerator("basicImage"),
+	"image-text-to-image": snippetGenerator("imageToImage", prepareImageToImageInput),
 	"image-text-to-text": snippetGenerator("conversational"),
+	"image-text-to-video": snippetGenerator("imageToVideo", prepareImageToImageInput),
 	"image-to-image": snippetGenerator("imageToImage", prepareImageToImageInput),
 	"image-to-text": snippetGenerator("basicImage"),
 	"image-to-video": snippetGenerator("imageToVideo", prepareImageToImageInput),

package/src/tasks/cv/imageTextToImage.ts

@@ -0,0 +1,22 @@
+import type { ImageTextToImageInput } from "@huggingface/tasks";
+import { resolveProvider } from "../../lib/getInferenceProviderMapping.js";
+import { getProviderHelper } from "../../lib/getProviderHelper.js";
+import type { BaseArgs, Options } from "../../types.js";
+import { innerRequest } from "../../utils/request.js";
+
+export type ImageTextToImageArgs = BaseArgs & ImageTextToImageInput;
+
+/**
+ * This task takes an image and text input and outputs a new generated image.
+ * Recommended model: black-forest-labs/FLUX.2-dev
+ */
+export async function imageTextToImage(args: ImageTextToImageArgs, options?: Options): Promise<Blob> {
+	const provider = await resolveProvider(args.provider, args.model, args.endpointUrl);
+	const providerHelper = getProviderHelper(provider, "image-text-to-image");
+	const payload = await providerHelper.preparePayloadAsync(args);
+	const { data: res, requestContext } = await innerRequest<Blob>(payload, providerHelper, {
+		...options,
+		task: "image-text-to-image",
+	});
+	return providerHelper.getResponse(res, requestContext.url, requestContext.info.headers as Record<string, string>);
+}

package/src/tasks/cv/imageTextToVideo.ts

@@ -0,0 +1,22 @@
+import type { ImageTextToVideoInput } from "@huggingface/tasks";
+import { resolveProvider } from "../../lib/getInferenceProviderMapping.js";
+import { getProviderHelper } from "../../lib/getProviderHelper.js";
+import type { BaseArgs, Options } from "../../types.js";
+import { innerRequest } from "../../utils/request.js";
+
+export type ImageTextToVideoArgs = BaseArgs & ImageTextToVideoInput;
+
+/**
+ * This task takes an image and text input and outputs a generated video.
+ * Recommended model: Lightricks/LTX-Video
+ */
+export async function imageTextToVideo(args: ImageTextToVideoArgs, options?: Options): Promise<Blob> {
+	const provider = await resolveProvider(args.provider, args.model, args.endpointUrl);
+	const providerHelper = getProviderHelper(provider, "image-text-to-video");
+	const payload = await providerHelper.preparePayloadAsync(args);
+	const { data: res, requestContext } = await innerRequest<Blob>(payload, providerHelper, {
+		...options,
+		task: "image-text-to-video",
+	});
+	return providerHelper.getResponse(res, requestContext.url, requestContext.info.headers as Record<string, string>);
+}

package/src/tasks/index.ts

@@ -14,6 +14,8 @@ export * from "./cv/imageSegmentation.js";
 export * from "./cv/imageToImage.js";
 export * from "./cv/imageToText.js";
 export * from "./cv/imageToVideo.js";
+export * from "./cv/imageTextToImage.js";
+export * from "./cv/imageTextToVideo.js";
 export * from "./cv/objectDetection.js";
 export * from "./cv/textToImage.js";
 export * from "./cv/textToVideo.js";