@huggingface/tasks 0.16.4 → 0.16.6

package/dist/esm/tasks/depth-estimation/data.js

@@ -39,7 +39,7 @@ const taskData = {
         },
         {
             description: "A robust depth estimation model.",
-            id: "apple/DepthPro",
+            id: "apple/DepthPro-hf",
         },
     ],
     spaces: [

package/dist/esm/tasks/image-text-to-text/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/image-text-to-text/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAyGf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/image-text-to-text/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cA6Gf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/image-text-to-text/data.js

@@ -45,7 +45,7 @@ const taskData = {
         },
         {
             description: "A screenshot understanding model used to control computers.",
-            id: "showlab/ShowUI-2B",
+            id: "microsoft/OmniParser-v2.0",
         },
         {
             description: "Cutting-edge vision language model.",
@@ -60,12 +60,16 @@ const taskData = {
             id: "Qwen/Qwen2.5-VL-7B-Instruct",
         },
         {
-            description: "Image-text-to-text model with reasoning capabilities.",
-            id: "Qwen/QVQ-72B-Preview",
+            description: "Image-text-to-text model with agentic capabilities.",
+            id: "microsoft/Magma-8B",
         },
         {
             description: "Strong image-text-to-text model focused on documents.",
-            id: "stepfun-ai/GOT-OCR2_0",
+            id: "allenai/olmOCR-7B-0225-preview",
+        },
+        {
+            description: "Small yet strong image-text-to-text model.",
+            id: "ibm-granite/granite-vision-3.2-2b",
         },
     ],
     spaces: [
@@ -82,8 +86,8 @@ const taskData = {
             id: "akhaliq/Molmo-7B-D-0924",
         },
         {
-            description: "An image-text-to-text application focused on documents.",
-            id: "stepfun-ai/GOT_official_online_demo",
+            description: "Powerful vision language assistant that can understand multiple images.",
+            id: "HuggingFaceTB/SmolVLM2",
         },
         {
             description: "An application for chatting with an image-text-to-text model.",

package/dist/esm/tasks/keypoint-detection/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/keypoint-detection/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAiDf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/keypoint-detection/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAqDf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/keypoint-detection/data.js

@@ -25,6 +25,10 @@ const taskData = {
             description: "A robust keypoint detection model.",
             id: "magic-leap-community/superpoint",
         },
+        {
+            description: "A robust keypoint matching model.",
+            id: "magic-leap-community/superglue_outdoor",
+        },
         {
             description: "Strong keypoint detection model used to detect human pose.",
             id: "facebook/sapiens-pose-1b",

package/dist/esm/tasks/object-detection/data.js

@@ -43,12 +43,12 @@ const taskData = {
             id: "facebook/detr-resnet-50",
         },
         {
-            description: "Real-time and accurate object detection model.",
-            id: "jameslahm/yolov10x",
+            description: "Accurate object detection model.",
+            id: "IDEA-Research/dab-detr-resnet-50",
         },
         {
-            description: "Fast and accurate object detection model trained on COCO and Object365 datasets.",
-            id: "PekingU/rtdetr_r18vd_coco_o365",
+            description: "Fast and accurate object detection model.",
+            id: "PekingU/rtdetr_v2_r50vd",
         },
         {
             description: "Object detection model for low-lying objects.",
@@ -66,7 +66,7 @@ const taskData = {
         },
         {
             description: "A cutting-edge object detection application.",
-            id: "Ultralytics/YOLO11",
+            id: "sunsmarterjieleaf/yolov12",
         },
         {
             description: "An object tracking, segmentation and inpainting application.",

package/dist/esm/tasks/text-generation/data.js

@@ -71,7 +71,7 @@ const taskData = {
         },
         {
             description: "A very powerful model with reasoning capabilities.",
-            id: "PowerInfer/SmallThinker-3B-Preview",
+            id: "simplescaling/s1.1-32B",
         },
         {
             description: "Strong conversational model that supports very long instructions.",

package/dist/esm/tasks/text-to-speech/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/text-to-speech/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAiFf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/text-to-speech/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAqFf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/text-to-speech/data.js

@@ -74,6 +74,10 @@ const taskData = {
             description: "An application that synthesizes emotional speech for diverse speaker prompts.",
             id: "parler-tts/parler-tts-expresso",
         },
+        {
+            description: "An application that generates podcast episodes.",
+            id: "ngxson/kokoro-podcast-generator",
+        },
     ],
     summary: "Text-to-Speech (TTS) is the task of generating natural sounding speech given text input. TTS models can be extended to have a single model that generates speech for multiple speakers and multiple languages.",
     widgetModels: ["suno/bark"],

package/dist/esm/tasks/text-to-video/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/text-to-video/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAiGf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/text-to-video/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAqGf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/text-to-video/data.js

@@ -71,6 +71,10 @@ const taskData = {
             description: "A text-to-video model focusing on physics-aware applications like robotics.",
             id: "nvidia/Cosmos-1.0-Diffusion-7B-Text2World",
         },
+        {
+            description: "A robust model for video generation.",
+            id: "Wan-AI/Wan2.1-T2V-1.3B",
+        },
     ],
     spaces: [
         {
@@ -79,7 +83,7 @@ const taskData = {
         },
         {
             description: "Consistent video generation application.",
-            id: "TIGER-Lab/T2V-Turbo-V2",
+            id: "Wan-AI/Wan2.1",
         },
         {
             description: "A cutting edge video generation application.",
@@ -87,7 +91,7 @@ const taskData = {
         },
     ],
     summary: "Text-to-video models can be used in any application that requires generating consistent sequence of images from text. ",
-    widgetModels: [],
+    widgetModels: ["tencent/HunyuanVideo"],
     youtubeId: undefined,
 };
 export default taskData;

package/dist/esm/tasks/video-text-to-text/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/video-text-to-text/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cA6Df,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/video-text-to-text/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAqEf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/video-text-to-text/data.js

@@ -43,6 +43,10 @@ const taskData = {
             description: "Strong video-text-to-text model with reasoning capabilities.",
             id: "GoodiesHere/Apollo-LMMs-Apollo-7B-t32",
         },
+        {
+            description: "Strong video-text-to-text model.",
+            id: "HuggingFaceTB/SmolVLM2-2.2B-Instruct",
+        },
     ],
     spaces: [
         {
@@ -53,6 +57,10 @@ const taskData = {
             description: "A leaderboard for various video-text-to-text models.",
             id: "opencompass/openvlm_video_leaderboard",
         },
+        {
+            description: "An application to generate highlights from a video.",
+            id: "HuggingFaceTB/SmolVLM2-HighlightGenerator",
+        },
     ],
     summary: "Video-text-to-text models take in a video and a text prompt and output text. These models are also called video-language models.",
     widgetModels: [""],

package/dist/esm/tasks/zero-shot-classification/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/zero-shot-classification/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAiEf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/zero-shot-classification/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAqEf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/zero-shot-classification/data.js

@@ -56,6 +56,10 @@ const taskData = {
             description: "Cutting-edge zero-shot multilingual text classification model.",
             id: "MoritzLaurer/ModernBERT-large-zeroshot-v2.0",
         },
+        {
+            description: "Zero-shot text classification model that can be used for topic and sentiment classification.",
+            id: "knowledgator/gliclass-modern-base-v2.0-init",
+        },
     ],
     spaces: [],
     summary: "Zero-shot text classification is a task in natural language processing where a model is trained on a set of labeled examples but is then able to classify new examples from previously unseen classes.",

package/dist/esm/tasks/zero-shot-image-classification/data.js

@@ -51,11 +51,11 @@ const taskData = {
         },
         {
             description: "Strong zero-shot image classification model.",
-            id: "google/siglip-so400m-patch14-224",
+            id: "google/siglip2-base-patch16-224",
         },
         {
             description: "Robust zero-shot image classification model.",
-            id: "microsoft/LLM2CLIP-EVA02-L-14-336",
+            id: "intfloat/mmE5-mllama-11b-instruct",
         },
         {
             description: "Powerful zero-shot image classification model supporting 94 languages.",

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@huggingface/tasks",
   "packageManager": "pnpm@8.10.5",
-  "version": "0.16.4",
+  "version": "0.16.6",
   "description": "List of ML tasks for huggingface.co/tasks",
   "repository": "https://github.com/huggingface/huggingface.js.git",
   "publishConfig": {

package/src/model-libraries.ts

@@ -394,6 +394,12 @@ export const MODEL_LIBRARIES_UI_ELEMENTS = {
 		repoUrl: "https://github.com/Tencent/HunyuanDiT",
 		countDownloads: `path:"pytorch_model_ema.pt" OR path:"pytorch_model_distill.pt"`,
 	},
+	"hunyuan3d-2": {
+		prettyLabel: "Hunyuan3D-2",
+		repoName: "Hunyuan3D-2",
+		repoUrl: "https://github.com/Tencent/Hunyuan3D-2",
+		countDownloads: `path:"model_index.json" OR path:"config.yaml"`,
+	},
 	imstoucan: {
 		prettyLabel: "IMS Toucan",
 		repoName: "IMS-Toucan",

package/src/snippets/inputs.ts

@@ -96,6 +96,8 @@ const inputsAudioClassification = () => `"sample1.flac"`;
 
 const inputsTextToImage = () => `"Astronaut riding a horse"`;
 
+const inputsTextToVideo = () => `"A young man walking on the street"`;
+
 const inputsTextToSpeech = () => `"The answer to the universe is 42"`;
 
 const inputsTextToAudio = () => `"liquid drum and bass, atmospheric synths, airy sounds"`;
@@ -130,6 +132,7 @@ const modelInputSnippets: {
 	"text-generation": inputsTextGeneration,
 	"image-text-to-text": inputsTextGeneration,
 	"text-to-image": inputsTextToImage,
+	"text-to-video": inputsTextToVideo,
 	"text-to-speech": inputsTextToSpeech,
 	"text-to-audio": inputsTextToAudio,
 	"text2text-generation": inputsText2TextGeneration,

package/src/snippets/js.ts

@@ -275,6 +275,33 @@ query({"inputs": ${getModelInputSnippet(model)}}).then((response) => {
 	];
 };
 
+export const snippetTextToVideo = (
+	model: ModelDataMinimal,
+	accessToken: string,
+	provider: SnippetInferenceProvider
+): InferenceSnippet[] => {
+	return ["fal-ai", "replicate"].includes(provider)
+		? [
+				{
+					client: "huggingface.js",
+					content: `\
+import { HfInference } from "@huggingface/inference";
+
+const client = new HfInference("${accessToken || `{API_TOKEN}`}");
+
+const video = await client.textToVideo({
+	model: "${model.id}",
+	provider: "${provider}",
+	inputs: ${getModelInputSnippet(model)},
+	parameters: { num_inference_steps: 5 },
+});
+// Use the generated video (it's a Blob)
+`,
+				},
+		  ]
+		: [];
+};
+
 export const snippetTextToAudio = (
 	model: ModelDataMinimal,
 	accessToken: string,
@@ -420,6 +447,7 @@ export const jsSnippets: Partial<
 	"sentence-similarity": snippetBasic,
 	"automatic-speech-recognition": snippetAutomaticSpeechRecognition,
 	"text-to-image": snippetTextToImage,
+	"text-to-video": snippetTextToVideo,
 	"text-to-speech": snippetTextToAudio,
 	"text-to-audio": snippetTextToAudio,
 	"audio-to-audio": snippetFile,

package/src/snippets/python.ts

@@ -308,6 +308,27 @@ image = Image.open(io.BytesIO(image_bytes))`,
 	];
 };
 
+export const snippetTextToVideo = (
+	model: ModelDataMinimal,
+	accessToken: string,
+	provider: SnippetInferenceProvider
+): InferenceSnippet[] => {
+	return ["fal-ai", "replicate"].includes(provider)
+		? [
+				{
+					client: "huggingface_hub",
+					content: `\
+${snippetImportInferenceClient(accessToken, provider)}
+
+video = client.text_to_video(
+	${getModelInputSnippet(model)},
+	model="${model.id}"
+)`,
+				},
+		  ]
+		: [];
+};
+
 export const snippetTabular = (model: ModelDataMinimal): InferenceSnippet[] => {
 	return [
 		{
@@ -412,6 +433,7 @@ export const pythonSnippets: Partial<
 	"sentence-similarity": snippetBasic,
 	"automatic-speech-recognition": snippetFile,
 	"text-to-image": snippetTextToImage,
+	"text-to-video": snippetTextToVideo,
 	"text-to-speech": snippetTextToAudio,
 	"text-to-audio": snippetTextToAudio,
 	"audio-to-audio": snippetFile,

package/src/tasks/depth-estimation/data.ts

@@ -41,7 +41,7 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "A robust depth estimation model.",
-			id: "apple/DepthPro",
+			id: "apple/DepthPro-hf",
 		},
 	],
 	spaces: [

package/src/tasks/image-text-to-text/data.ts

@@ -48,7 +48,7 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "A screenshot understanding model used to control computers.",
-			id: "showlab/ShowUI-2B",
+			id: "microsoft/OmniParser-v2.0",
 		},
 		{
 			description: "Cutting-edge vision language model.",
@@ -63,12 +63,16 @@ const taskData: TaskDataCustom = {
 			id: "Qwen/Qwen2.5-VL-7B-Instruct",
 		},
 		{
-			description: "Image-text-to-text model with reasoning capabilities.",
-			id: "Qwen/QVQ-72B-Preview",
+			description: "Image-text-to-text model with agentic capabilities.",
+			id: "microsoft/Magma-8B",
 		},
 		{
 			description: "Strong image-text-to-text model focused on documents.",
-			id: "stepfun-ai/GOT-OCR2_0",
+			id: "allenai/olmOCR-7B-0225-preview",
+		},
+		{
+			description: "Small yet strong image-text-to-text model.",
+			id: "ibm-granite/granite-vision-3.2-2b",
 		},
 	],
 	spaces: [
@@ -85,8 +89,8 @@ const taskData: TaskDataCustom = {
 			id: "akhaliq/Molmo-7B-D-0924",
 		},
 		{
-			description: "An image-text-to-text application focused on documents.",
-			id: "stepfun-ai/GOT_official_online_demo",
+			description: "Powerful vision language assistant that can understand multiple images.",
+			id: "HuggingFaceTB/SmolVLM2",
 		},
 		{
 			description: "An application for chatting with an image-text-to-text model.",

package/src/tasks/keypoint-detection/data.ts

@@ -27,6 +27,10 @@ const taskData: TaskDataCustom = {
 			description: "A robust keypoint detection model.",
 			id: "magic-leap-community/superpoint",
 		},
+		{
+			description: "A robust keypoint matching model.",
+			id: "magic-leap-community/superglue_outdoor",
+		},
 		{
 			description: "Strong keypoint detection model used to detect human pose.",
 			id: "facebook/sapiens-pose-1b",

package/src/tasks/object-detection/data.ts

@@ -47,12 +47,12 @@ const taskData: TaskDataCustom = {
 			id: "facebook/detr-resnet-50",
 		},
 		{
-			description: "Real-time and accurate object detection model.",
-			id: "jameslahm/yolov10x",
+			description: "Accurate object detection model.",
+			id: "IDEA-Research/dab-detr-resnet-50",
 		},
 		{
-			description: "Fast and accurate object detection model trained on COCO and Object365 datasets.",
-			id: "PekingU/rtdetr_r18vd_coco_o365",
+			description: "Fast and accurate object detection model.",
+			id: "PekingU/rtdetr_v2_r50vd",
 		},
 		{
 			description: "Object detection model for low-lying objects.",
@@ -70,7 +70,7 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "A cutting-edge object detection application.",
-			id: "Ultralytics/YOLO11",
+			id: "sunsmarterjieleaf/yolov12",
 		},
 		{
 			description: "An object tracking, segmentation and inpainting application.",

package/src/tasks/text-generation/data.ts

@@ -76,7 +76,7 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "A very powerful model with reasoning capabilities.",
-			id: "PowerInfer/SmallThinker-3B-Preview",
+			id: "simplescaling/s1.1-32B",
 		},
 		{
 			description: "Strong conversational model that supports very long instructions.",

package/src/tasks/text-to-speech/data.ts

@@ -76,6 +76,10 @@ const taskData: TaskDataCustom = {
 			description: "An application that synthesizes emotional speech for diverse speaker prompts.",
 			id: "parler-tts/parler-tts-expresso",
 		},
+		{
+			description: "An application that generates podcast episodes.",
+			id: "ngxson/kokoro-podcast-generator",
+		},
 	],
 	summary:
 		"Text-to-Speech (TTS) is the task of generating natural sounding speech given text input. TTS models can be extended to have a single model that generates speech for multiple speakers and multiple languages.",

package/src/tasks/text-to-video/data.ts

@@ -78,6 +78,10 @@ const taskData: TaskDataCustom = {
 			description: "A text-to-video model focusing on physics-aware applications like robotics.",
 			id: "nvidia/Cosmos-1.0-Diffusion-7B-Text2World",
 		},
+		{
+			description: "A robust model for video generation.",
+			id: "Wan-AI/Wan2.1-T2V-1.3B",
+		},
 	],
 	spaces: [
 		{
@@ -86,7 +90,7 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "Consistent video generation application.",
-			id: "TIGER-Lab/T2V-Turbo-V2",
+			id: "Wan-AI/Wan2.1",
 		},
 		{
 			description: "A cutting edge video generation application.",
@@ -95,7 +99,7 @@ const taskData: TaskDataCustom = {
 	],
 	summary:
 		"Text-to-video models can be used in any application that requires generating consistent sequence of images from text. ",
-	widgetModels: [],
+	widgetModels: ["tencent/HunyuanVideo"],
 	youtubeId: undefined,
 };
 

package/src/tasks/video-text-to-text/data.ts

@@ -46,6 +46,10 @@ const taskData: TaskDataCustom = {
 			description: "Strong video-text-to-text model with reasoning capabilities.",
 			id: "GoodiesHere/Apollo-LMMs-Apollo-7B-t32",
 		},
+		{
+			description: "Strong video-text-to-text model.",
+			id: "HuggingFaceTB/SmolVLM2-2.2B-Instruct",
+		},
 	],
 	spaces: [
 		{
@@ -56,6 +60,10 @@ const taskData: TaskDataCustom = {
 			description: "A leaderboard for various video-text-to-text models.",
 			id: "opencompass/openvlm_video_leaderboard",
 		},
+		{
+			description: "An application to generate highlights from a video.",
+			id: "HuggingFaceTB/SmolVLM2-HighlightGenerator",
+		},
 	],
 	summary:
 		"Video-text-to-text models take in a video and a text prompt and output text. These models are also called video-language models.",

package/src/tasks/zero-shot-classification/data.ts

@@ -60,6 +60,10 @@ const taskData: TaskDataCustom = {
 			description: "Cutting-edge zero-shot multilingual text classification model.",
 			id: "MoritzLaurer/ModernBERT-large-zeroshot-v2.0",
 		},
+		{
+			description: "Zero-shot text classification model that can be used for topic and sentiment classification.",
+			id: "knowledgator/gliclass-modern-base-v2.0-init",
+		},
 	],
 	spaces: [],
 	summary:

package/src/tasks/zero-shot-image-classification/data.ts

@@ -53,11 +53,11 @@ const taskData: TaskDataCustom = {
 		},
 		{
 			description: "Strong zero-shot image classification model.",
-			id: "google/siglip-so400m-patch14-224",
+			id: "google/siglip2-base-patch16-224",
 		},
 		{
 			description: "Robust zero-shot image classification model.",
-			id: "microsoft/LLM2CLIP-EVA02-L-14-336",
+			id: "intfloat/mmE5-mllama-11b-instruct",
 		},
 		{
 			description: "Powerful zero-shot image classification model supporting 94 languages.",