@huggingface/inference 3.15.0 → 4.0.0

package/dist/commonjs/providers/ovhcloud.js

@@ -18,8 +18,8 @@
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.OvhCloudTextGenerationTask = exports.OvhCloudConversationalTask = void 0;
 const providerHelper_js_1 = require("./providerHelper.js");
-const InferenceOutputError_js_1 = require("../lib/InferenceOutputError.js");
 const omit_js_1 = require("../utils/omit.js");
+const errors_js_1 = require("../errors.js");
 const OVHCLOUD_API_BASE_URL = "https://oai.endpoints.kepler.ai.cloud.ovh.net";
 class OvhCloudConversationalTask extends providerHelper_js_1.BaseConversationalTask {
     constructor() {
@@ -54,7 +54,7 @@ class OvhCloudTextGenerationTask extends providerHelper_js_1.BaseTextGenerationT
                 generated_text: completion.text,
             };
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected OVHcloud text generation response format");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from OVHcloud text generation API");
     }
 }
 exports.OvhCloudTextGenerationTask = OvhCloudTextGenerationTask;

package/dist/commonjs/providers/providerHelper.js

@@ -2,7 +2,7 @@
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.BaseTextGenerationTask = exports.BaseConversationalTask = exports.TaskProviderHelper = void 0;
 const config_js_1 = require("../config.js");
-const InferenceOutputError_js_1 = require("../lib/InferenceOutputError.js");
+const errors_js_1 = require("../errors.js");
 const toArray_js_1 = require("../utils/toArray.js");
 /**
  * Base class for task-specific provider helpers
@@ -78,7 +78,7 @@ class BaseConversationalTask extends TaskProviderHelper {
             typeof response?.usage === "object") {
             return response;
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected ChatCompletionOutput");
+        throw new errors_js_1.InferenceClientProviderOutputError("Expected ChatCompletionOutput");
     }
 }
 exports.BaseConversationalTask = BaseConversationalTask;
@@ -102,7 +102,7 @@ class BaseTextGenerationTask extends TaskProviderHelper {
             res.every((x) => typeof x === "object" && !!x && "generated_text" in x && typeof x.generated_text === "string")) {
             return res[0];
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Array<{generated_text: string}>");
+        throw new errors_js_1.InferenceClientProviderOutputError("Expected Array<{generated_text: string}>");
     }
 }
 exports.BaseTextGenerationTask = BaseTextGenerationTask;

package/dist/commonjs/providers/replicate.js

@@ -17,7 +17,7 @@ exports.ReplicateTextToVideoTask = exports.ReplicateTextToSpeechTask = exports.R
  *
  * Thanks!
  */
-const InferenceOutputError_js_1 = require("../lib/InferenceOutputError.js");
+const errors_js_1 = require("../errors.js");
 const isUrl_js_1 = require("../lib/isUrl.js");
 const omit_js_1 = require("../utils/omit.js");
 const providerHelper_js_1 = require("./providerHelper.js");
@@ -84,7 +84,7 @@ class ReplicateTextToImageTask extends ReplicateTask {
             const urlResponse = await fetch(res.output[0]);
             return await urlResponse.blob();
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Replicate text-to-image response format");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Replicate text-to-image API");
     }
 }
 exports.ReplicateTextToImageTask = ReplicateTextToImageTask;
@@ -115,7 +115,7 @@ class ReplicateTextToSpeechTask extends ReplicateTask {
                 }
             }
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Blob or object with output");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Replicate text-to-speech API");
     }
 }
 exports.ReplicateTextToSpeechTask = ReplicateTextToSpeechTask;
@@ -129,7 +129,7 @@ class ReplicateTextToVideoTask extends ReplicateTask {
             const urlResponse = await fetch(response.output);
             return await urlResponse.blob();
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected { output: string }");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Replicate text-to-video API");
     }
 }
 exports.ReplicateTextToVideoTask = ReplicateTextToVideoTask;

package/dist/commonjs/providers/sambanova.d.ts

@@ -1,3 +1,19 @@
+/**
+ * See the registered mapping of HF model ID => Sambanova model ID here:
+ *
+ * https://huggingface.co/api/partners/sambanova/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 Sambanova 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 Sambanova, please open an issue on the present repo
+ * and we will tag Sambanova team members.
+ *
+ * Thanks!
+ */
 import type { FeatureExtractionOutput } from "@huggingface/tasks";
 import type { BodyParams } from "../types.js";
 import type { FeatureExtractionTaskHelper } from "./providerHelper.js";

package/dist/commonjs/providers/sambanova.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"sambanova.d.ts","sourceRoot":"","sources":["../../../src/providers/sambanova.ts"],"names":[],"mappings":"AAkBA,OAAO,KAAK,EAAE,uBAAuB,EAAE,MAAM,oBAAoB,CAAC;AAClE,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAC9C,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,qBAAqB,CAAC;AACvE,OAAO,EAAE,sBAAsB,EAAE,kBAAkB,EAAE,MAAM,qBAAqB,CAAC;AAEjF,qBAAa,2BAA4B,SAAQ,sBAAsB;;CAItE;AAED,qBAAa,8BAA+B,SAAQ,kBAAmB,YAAW,2BAA2B;;IAKnG,SAAS,IAAI,MAAM;IAIb,WAAW,CAAC,QAAQ,EAAE,uBAAuB,GAAG,OAAO,CAAC,uBAAuB,CAAC;IAStF,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;CAOpE"}
+{"version":3,"file":"sambanova.d.ts","sourceRoot":"","sources":["../../../src/providers/sambanova.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,OAAO,KAAK,EAAE,uBAAuB,EAAE,MAAM,oBAAoB,CAAC;AAClE,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAC9C,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,qBAAqB,CAAC;AACvE,OAAO,EAAE,sBAAsB,EAAE,kBAAkB,EAAE,MAAM,qBAAqB,CAAC;AAGjF,qBAAa,2BAA4B,SAAQ,sBAAsB;;CAItE;AAED,qBAAa,8BAA+B,SAAQ,kBAAmB,YAAW,2BAA2B;;IAKnG,SAAS,IAAI,MAAM;IAIb,WAAW,CAAC,QAAQ,EAAE,uBAAuB,GAAG,OAAO,CAAC,uBAAuB,CAAC;IAStF,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;CAOpE"}

package/dist/commonjs/providers/sambanova.js

@@ -1,24 +1,8 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.SambanovaFeatureExtractionTask = exports.SambanovaConversationalTask = void 0;
-/**
- * See the registered mapping of HF model ID => Sambanova model ID here:
- *
- * https://huggingface.co/api/partners/sambanova/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 Sambanova 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 Sambanova, please open an issue on the present repo
- * and we will tag Sambanova team members.
- *
- * Thanks!
- */
-const InferenceOutputError_js_1 = require("../lib/InferenceOutputError.js");
 const providerHelper_js_1 = require("./providerHelper.js");
+const errors_js_1 = require("../errors.js");
 class SambanovaConversationalTask extends providerHelper_js_1.BaseConversationalTask {
     constructor() {
         super("sambanova", "https://api.sambanova.ai");
@@ -36,7 +20,7 @@ class SambanovaFeatureExtractionTask extends providerHelper_js_1.TaskProviderHel
         if (typeof response === "object" && "data" in response && Array.isArray(response.data)) {
             return response.data.map((item) => item.embedding);
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Sambanova feature-extraction (embeddings) response format to be {'data' : list of {'embedding' : number[]}}");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Sambanova feature-extraction (embeddings) API");
     }
     preparePayload(params) {
         return {

package/dist/commonjs/providers/together.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"together.d.ts","sourceRoot":"","sources":["../../../src/providers/together.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,OAAO,KAAK,EAAE,oBAAoB,EAAE,oBAAoB,EAAE,gCAAgC,EAAE,MAAM,oBAAoB,CAAC;AAEvH,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAE9C,OAAO,EACN,sBAAsB,EACtB,sBAAsB,EACtB,kBAAkB,EAClB,KAAK,qBAAqB,EAC1B,MAAM,qBAAqB,CAAC;AAI7B,UAAU,4BAA6B,SAAQ,IAAI,CAAC,oBAAoB,EAAE,SAAS,CAAC;IACnF,OAAO,EAAE,KAAK,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;QACb,aAAa,EAAE,gCAAgC,CAAC;QAChD,IAAI,EAAE,MAAM,CAAC;QACb,QAAQ,EAAE,OAAO,CAAC;QAClB,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;CACH;AAED,UAAU,6BAA6B;IACtC,IAAI,EAAE,KAAK,CAAC;QACX,QAAQ,EAAE,MAAM,CAAC;KACjB,CAAC,CAAC;CACH;AAED,qBAAa,0BAA2B,SAAQ,sBAAsB;;CAIrE;AAED,qBAAa,0BAA2B,SAAQ,sBAAsB;;IAK5D,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;IAQrD,WAAW,CAAC,QAAQ,EAAE,4BAA4B,GAAG,OAAO,CAAC,oBAAoB,CAAC;CAcjG;AAED,qBAAa,uBAAwB,SAAQ,kBAAmB,YAAW,qBAAqB;;IAK/F,SAAS,IAAI,MAAM;IAInB,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;IAUrD,WAAW,CAAC,QAAQ,EAAE,6BAA6B,EAAE,UAAU,CAAC,EAAE,KAAK,GAAG,MAAM,GAAG,OAAO,CAAC,MAAM,GAAG,IAAI,CAAC;CAkB/G"}
+{"version":3,"file":"together.d.ts","sourceRoot":"","sources":["../../../src/providers/together.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,OAAO,KAAK,EAAE,oBAAoB,EAAE,oBAAoB,EAAE,gCAAgC,EAAE,MAAM,oBAAoB,CAAC;AACvH,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAE9C,OAAO,EACN,sBAAsB,EACtB,sBAAsB,EACtB,kBAAkB,EAClB,KAAK,qBAAqB,EAC1B,MAAM,qBAAqB,CAAC;AAK7B,UAAU,4BAA6B,SAAQ,IAAI,CAAC,oBAAoB,EAAE,SAAS,CAAC;IACnF,OAAO,EAAE,KAAK,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;QACb,aAAa,EAAE,gCAAgC,CAAC;QAChD,IAAI,EAAE,MAAM,CAAC;QACb,QAAQ,EAAE,OAAO,CAAC;QAClB,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;CACH;AAED,UAAU,6BAA6B;IACtC,IAAI,EAAE,KAAK,CAAC;QACX,QAAQ,EAAE,MAAM,CAAC;KACjB,CAAC,CAAC;CACH;AAED,qBAAa,0BAA2B,SAAQ,sBAAsB;;CAIrE;AAED,qBAAa,0BAA2B,SAAQ,sBAAsB;;IAK5D,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;IAQrD,WAAW,CAAC,QAAQ,EAAE,4BAA4B,GAAG,OAAO,CAAC,oBAAoB,CAAC;CAcjG;AAED,qBAAa,uBAAwB,SAAQ,kBAAmB,YAAW,qBAAqB;;IAK/F,SAAS,IAAI,MAAM;IAInB,cAAc,CAAC,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;IAUrD,WAAW,CAAC,QAAQ,EAAE,6BAA6B,EAAE,UAAU,CAAC,EAAE,KAAK,GAAG,MAAM,GAAG,OAAO,CAAC,MAAM,GAAG,IAAI,CAAC;CAkB/G"}

package/dist/commonjs/providers/together.js

@@ -1,9 +1,9 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.TogetherTextToImageTask = exports.TogetherTextGenerationTask = exports.TogetherConversationalTask = void 0;
-const InferenceOutputError_js_1 = require("../lib/InferenceOutputError.js");
 const omit_js_1 = require("../utils/omit.js");
 const providerHelper_js_1 = require("./providerHelper.js");
+const errors_js_1 = require("../errors.js");
 const TOGETHER_API_BASE_URL = "https://api.together.xyz";
 class TogetherConversationalTask extends providerHelper_js_1.BaseConversationalTask {
     constructor() {
@@ -32,7 +32,7 @@ class TogetherTextGenerationTask extends providerHelper_js_1.BaseTextGenerationT
                 generated_text: completion.text,
             };
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Together text generation response format");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Together text generation API");
     }
 }
 exports.TogetherTextGenerationTask = TogetherTextGenerationTask;
@@ -65,7 +65,7 @@ class TogetherTextToImageTask extends providerHelper_js_1.TaskProviderHelper {
             }
             return fetch(`data:image/jpeg;base64,${base64Data}`).then((res) => res.blob());
         }
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected Together text-to-image response format");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from Together text-to-image API");
     }
 }
 exports.TogetherTextToImageTask = TogetherTextToImageTask;

package/dist/commonjs/tasks/audio/automaticSpeechRecognition.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"automaticSpeechRecognition.d.ts","sourceRoot":"","sources":["../../../../src/tasks/audio/automaticSpeechRecognition.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,+BAA+B,EAAE,gCAAgC,EAAE,MAAM,oBAAoB,CAAC;AAI5G,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,MAAM,gBAAgB,CAAC;AAExD,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAEnD,MAAM,MAAM,8BAA8B,GAAG,QAAQ,GAAG,CAAC,+BAA+B,GAAG,gBAAgB,CAAC,CAAC;AAC7G;;;GAGG;AACH,wBAAsB,0BAA0B,CAC/C,IAAI,EAAE,8BAA8B,EACpC,OAAO,CAAC,EAAE,OAAO,GACf,OAAO,CAAC,gCAAgC,CAAC,CAa3C"}
+{"version":3,"file":"automaticSpeechRecognition.d.ts","sourceRoot":"","sources":["../../../../src/tasks/audio/automaticSpeechRecognition.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,+BAA+B,EAAE,gCAAgC,EAAE,MAAM,oBAAoB,CAAC;AAG5G,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,MAAM,gBAAgB,CAAC;AAExD,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAGnD,MAAM,MAAM,8BAA8B,GAAG,QAAQ,GAAG,CAAC,+BAA+B,GAAG,gBAAgB,CAAC,CAAC;AAC7G;;;GAGG;AACH,wBAAsB,0BAA0B,CAC/C,IAAI,EAAE,8BAA8B,EACpC,OAAO,CAAC,EAAE,OAAO,GACf,OAAO,CAAC,gCAAgC,CAAC,CAa3C"}

package/dist/commonjs/tasks/audio/automaticSpeechRecognition.js

@@ -3,8 +3,8 @@ Object.defineProperty(exports, "__esModule", { value: true });
 exports.automaticSpeechRecognition = automaticSpeechRecognition;
 const getInferenceProviderMapping_js_1 = require("../../lib/getInferenceProviderMapping.js");
 const getProviderHelper_js_1 = require("../../lib/getProviderHelper.js");
-const InferenceOutputError_js_1 = require("../../lib/InferenceOutputError.js");
 const request_js_1 = require("../../utils/request.js");
+const errors_js_1 = require("../../errors.js");
 /**
  * This task reads some audio input and outputs the said words within the audio files.
  * Recommended model (english language): facebook/wav2vec2-large-960h-lv60-self
@@ -19,7 +19,7 @@ async function automaticSpeechRecognition(args, options) {
     });
     const isValidOutput = typeof res?.text === "string";
     if (!isValidOutput) {
-        throw new InferenceOutputError_js_1.InferenceOutputError("Expected {text: string}");
+        throw new errors_js_1.InferenceClientProviderOutputError("Received malformed response from automatic-speech-recognition API");
     }
     return providerHelper.getResponse(res);
 }

package/dist/commonjs/utils/request.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"request.d.ts","sourceRoot":"","sources":["../../../src/utils/request.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,6BAA6B,CAAC;AAErE,OAAO,KAAK,EAAE,aAAa,EAAE,OAAO,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAIvE,MAAM,WAAW,eAAe,CAAC,CAAC;IACjC,IAAI,EAAE,CAAC,CAAC;IACR,cAAc,EAAE;QACf,GAAG,EAAE,MAAM,CAAC;QACZ,IAAI,EAAE,WAAW,CAAC;KAClB,CAAC;CACF;AAED;;GAEG;AACH,wBAAsB,YAAY,CAAC,CAAC,EACnC,IAAI,EAAE,WAAW,EACjB,cAAc,EAAE,UAAU,CAAC,OAAO,iBAAiB,CAAC,EACpD,OAAO,CAAC,EAAE,OAAO,GAAG;IACnB,oEAAoE;IACpE,IAAI,CAAC,EAAE,aAAa,CAAC;IACrB,oCAAoC;IACpC,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB,GACC,OAAO,CAAC,eAAe,CAAC,CAAC,CAAC,CAAC,CAoC7B;AAED;;GAEG;AACH,wBAAuB,qBAAqB,CAAC,CAAC,EAC7C,IAAI,EAAE,WAAW,EACjB,cAAc,EAAE,UAAU,CAAC,OAAO,iBAAiB,CAAC,EACpD,OAAO,CAAC,EAAE,OAAO,GAAG;IACnB,oEAAoE;IACpE,IAAI,CAAC,EAAE,aAAa,CAAC;IACrB,oCAAoC;IACpC,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB,GACC,cAAc,CAAC,CAAC,CAAC,CAuFnB"}
+{"version":3,"file":"request.d.ts","sourceRoot":"","sources":["../../../src/utils/request.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,6BAA6B,CAAC;AAErE,OAAO,KAAK,EAAE,aAAa,EAAE,OAAO,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAMvE,MAAM,WAAW,eAAe,CAAC,CAAC;IACjC,IAAI,EAAE,CAAC,CAAC;IACR,cAAc,EAAE;QACf,GAAG,EAAE,MAAM,CAAC;QACZ,IAAI,EAAE,WAAW,CAAC;KAClB,CAAC;CACF;AAaD;;GAEG;AACH,wBAAsB,YAAY,CAAC,CAAC,EACnC,IAAI,EAAE,WAAW,EACjB,cAAc,EAAE,UAAU,CAAC,OAAO,iBAAiB,CAAC,EACpD,OAAO,CAAC,EAAE,OAAO,GAAG;IACnB,oEAAoE;IACpE,IAAI,CAAC,EAAE,aAAa,CAAC;IACrB,oCAAoC;IACpC,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB,GACC,OAAO,CAAC,eAAe,CAAC,CAAC,CAAC,CAAC,CAwE7B;AAED;;GAEG;AACH,wBAAuB,qBAAqB,CAAC,CAAC,EAC7C,IAAI,EAAE,WAAW,EACjB,cAAc,EAAE,UAAU,CAAC,OAAO,iBAAiB,CAAC,EACpD,OAAO,CAAC,EAAE,OAAO,GAAG;IACnB,oEAAoE;IACpE,IAAI,CAAC,EAAE,aAAa,CAAC;IACrB,oCAAoC;IACpC,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB,GACC,cAAc,CAAC,CAAC,CAAC,CAuJnB"}

package/dist/commonjs/utils/request.js

@@ -4,6 +4,15 @@ exports.innerRequest = innerRequest;
 exports.innerStreamingRequest = innerStreamingRequest;
 const makeRequestOptions_js_1 = require("../lib/makeRequestOptions.js");
 const parse_js_1 = require("../vendor/fetch-event-source/parse.js");
+const errors_js_1 = require("../errors.js");
+function requestArgsToJson(args) {
+    // Convert the entire args object to a JSON-serializable format
+    const argsWithData = args;
+    return JSON.parse(JSON.stringify({
+        ...argsWithData,
+        data: argsWithData.data ? "[Blob or ArrayBuffer]" : null,
+    }));
+}
 /**
  * Primitive to make custom calls to the inference provider
  */
@@ -19,17 +28,37 @@ async function innerRequest(args, providerHelper, options) {
         if (["application/json", "application/problem+json"].some((ct) => contentType?.startsWith(ct))) {
             const output = await response.json();
             if ([400, 422, 404, 500].includes(response.status) && options?.chatCompletion) {
-                throw new Error(`Server ${args.model} does not seem to support chat completion. Error: ${JSON.stringify(output.error)}`);
+                throw new errors_js_1.InferenceClientProviderApiError(`Provider ${args.provider} does not seem to support chat completion for model ${args.model} . Error: ${JSON.stringify(output.error)}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
-            if (output.error || output.detail) {
-                throw new Error(JSON.stringify(output.error ?? output.detail));
+            if (typeof output.error === "string" || typeof output.detail === "string") {
+                throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: ${output.error ?? output.detail}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
             else {
-                throw new Error(output);
+                throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: an HTTP error occurred when requesting the provider.`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
         }
         const message = contentType?.startsWith("text/plain;") ? await response.text() : undefined;
-        throw new Error(message ?? "An error occurred while fetching the blob");
+        throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: ${message ?? "an HTTP error occurred when requesting the provider"}`, {
+            url,
+            method: info.method ?? "GET",
+            headers: info.headers,
+            body: requestArgsToJson(args),
+        }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: message ?? "" });
     }
     if (response.headers.get("Content-Type")?.startsWith("application/json")) {
         const data = (await response.json());
@@ -51,24 +80,55 @@ async function* innerStreamingRequest(args, providerHelper, options) {
         if (response.headers.get("Content-Type")?.startsWith("application/json")) {
             const output = await response.json();
             if ([400, 422, 404, 500].includes(response.status) && options?.chatCompletion) {
-                throw new Error(`Server ${args.model} does not seem to support chat completion. Error: ${output.error}`);
+                throw new errors_js_1.InferenceClientProviderApiError(`Provider ${args.provider} does not seem to support chat completion for model ${args.model} . Error: ${JSON.stringify(output.error)}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
             if (typeof output.error === "string") {
-                throw new Error(output.error);
+                throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: ${output.error}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
             if (output.error && "message" in output.error && typeof output.error.message === "string") {
                 /// OpenAI errors
-                throw new Error(output.error.message);
+                throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: ${output.error.message}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
             // Sambanova errors
             if (typeof output.message === "string") {
-                throw new Error(output.message);
+                throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: ${output.message}`, {
+                    url,
+                    method: info.method ?? "GET",
+                    headers: info.headers,
+                    body: requestArgsToJson(args),
+                }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: output });
             }
         }
-        throw new Error(`Server response contains error: ${response.status}`);
+        throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: an HTTP error occurred when requesting the provider.`, {
+            url,
+            method: info.method ?? "GET",
+            headers: info.headers,
+            body: requestArgsToJson(args),
+        }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: "" });
     }
     if (!response.headers.get("content-type")?.startsWith("text/event-stream")) {
-        throw new Error(`Server does not support event stream content type, it returned ` + response.headers.get("content-type"));
+        throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: server does not support event stream content type, it returned ` +
+            response.headers.get("content-type"), {
+            url,
+            method: info.method ?? "GET",
+            headers: info.headers,
+            body: requestArgsToJson(args),
+        }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: "" });
     }
     if (!response.body) {
         return;
@@ -102,7 +162,12 @@ async function* innerStreamingRequest(args, providerHelper, options) {
                                 typeof data.error.message === "string"
                                 ? data.error.message
                                 : JSON.stringify(data.error);
-                        throw new Error(`Error forwarded from backend: ` + errorStr);
+                        throw new errors_js_1.InferenceClientProviderApiError(`Failed to perform inference: an occurred while streaming the response: ${errorStr}`, {
+                            url,
+                            method: info.method ?? "GET",
+                            headers: info.headers,
+                            body: requestArgsToJson(args),
+                        }, { requestId: response.headers.get("x-request-id") ?? "", status: response.status, body: data });
                     }
                     yield data;
                 }

package/dist/commonjs/vendor/type-fest/basic.d.ts

@@ -0,0 +1,33 @@
+/**
+Matches a JSON object.
+
+This type can be useful to enforce some input to be JSON-compatible or as a super-type to be extended from. Don't use this as a direct return type as the user would have to double-cast it: `jsonObject as unknown as CustomResponse`. Instead, you could extend your CustomResponse type from it to ensure your type only uses JSON-compatible types: `interface CustomResponse extends JsonObject { … }`.
+
+@category JSON
+*/
+export type JsonObject = {
+    [Key in string]: JsonValue;
+} & {
+    [Key in string]?: JsonValue | undefined;
+};
+/**
+Matches a JSON array.
+
+@category JSON
+*/
+export type JsonArray = JsonValue[] | readonly JsonValue[];
+/**
+Matches any valid JSON primitive value.
+
+@category JSON
+*/
+export type JsonPrimitive = string | number | boolean | null;
+/**
+Matches any valid JSON value.
+
+@see `Jsonify` if you need to transform a type to one that is assignable to `JsonValue`.
+
+@category JSON
+*/
+export type JsonValue = JsonPrimitive | JsonObject | JsonArray;
+//# sourceMappingURL=basic.d.ts.map

package/dist/commonjs/vendor/type-fest/basic.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"basic.d.ts","sourceRoot":"","sources":["../../../../src/vendor/type-fest/basic.ts"],"names":[],"mappings":"AAAA;;;;;;EAME;AACF,MAAM,MAAM,UAAU,GAAG;KAAG,GAAG,IAAI,MAAM,GAAG,SAAS;CAAE,GAAG;KAAG,GAAG,IAAI,MAAM,CAAC,CAAC,EAAE,SAAS,GAAG,SAAS;CAAE,CAAC;AAEtG;;;;EAIE;AACF,MAAM,MAAM,SAAS,GAAG,SAAS,EAAE,GAAG,SAAS,SAAS,EAAE,CAAC;AAE3D;;;;EAIE;AACF,MAAM,MAAM,aAAa,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,GAAG,IAAI,CAAC;AAE7D;;;;;;EAME;AACF,MAAM,MAAM,SAAS,GAAG,aAAa,GAAG,UAAU,GAAG,SAAS,CAAC"}

package/dist/commonjs/vendor/type-fest/basic.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/esm/errors.d.ts

@@ -0,0 +1,46 @@
+import type { JsonObject } from "./vendor/type-fest/basic.js";
+/**
+ * Base class for all inference-related errors.
+ */
+export declare abstract class InferenceClientError extends Error {
+    constructor(message: string);
+}
+export declare class InferenceClientInputError extends InferenceClientError {
+    constructor(message: string);
+}
+interface HttpRequest {
+    url: string;
+    method: string;
+    headers?: Record<string, string>;
+    body?: JsonObject;
+}
+interface HttpResponse {
+    requestId: string;
+    status: number;
+    body: JsonObject | string;
+}
+declare abstract class InferenceClientHttpRequestError extends InferenceClientError {
+    httpRequest: HttpRequest;
+    httpResponse: HttpResponse;
+    constructor(message: string, httpRequest: HttpRequest, httpResponse: HttpResponse);
+}
+/**
+ * Thrown when the HTTP request to the provider fails, e.g. due to API issues or server errors.
+ */
+export declare class InferenceClientProviderApiError extends InferenceClientHttpRequestError {
+    constructor(message: string, httpRequest: HttpRequest, httpResponse: HttpResponse);
+}
+/**
+ * Thrown when the HTTP request to the hub fails, e.g. due to API issues or server errors.
+ */
+export declare class InferenceClientHubApiError extends InferenceClientHttpRequestError {
+    constructor(message: string, httpRequest: HttpRequest, httpResponse: HttpResponse);
+}
+/**
+ * Thrown when the inference output returned by the provider is invalid / does not match the expectations
+ */
+export declare class InferenceClientProviderOutputError extends InferenceClientError {
+    constructor(message: string);
+}
+export {};
+//# sourceMappingURL=errors.d.ts.map

package/dist/esm/errors.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"errors.d.ts","sourceRoot":"","sources":["../../src/errors.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,6BAA6B,CAAC;AAE9D;;GAEG;AACH,8BAAsB,oBAAqB,SAAQ,KAAK;gBAC3C,OAAO,EAAE,MAAM;CAI3B;AAED,qBAAa,yBAA0B,SAAQ,oBAAoB;gBACtD,OAAO,EAAE,MAAM;CAI3B;AAED,UAAU,WAAW;IACpB,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACjC,IAAI,CAAC,EAAE,UAAU,CAAC;CAClB;AAED,UAAU,YAAY;IACrB,SAAS,EAAE,MAAM,CAAC;IAClB,MAAM,EAAE,MAAM,CAAC;IACf,IAAI,EAAE,UAAU,GAAG,MAAM,CAAC;CAC1B;AAED,uBAAe,+BAAgC,SAAQ,oBAAoB;IAC1E,WAAW,EAAE,WAAW,CAAC;IACzB,YAAY,EAAE,YAAY,CAAC;gBACf,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,WAAW,EAAE,YAAY,EAAE,YAAY;CAgBjF;AAED;;GAEG;AACH,qBAAa,+BAAgC,SAAQ,+BAA+B;gBACvE,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,WAAW,EAAE,YAAY,EAAE,YAAY;CAIjF;AAED;;GAEG;AACH,qBAAa,0BAA2B,SAAQ,+BAA+B;gBAClE,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,WAAW,EAAE,YAAY,EAAE,YAAY;CAIjF;AAED;;GAEG;AACH,qBAAa,kCAAmC,SAAQ,oBAAoB;gBAC/D,OAAO,EAAE,MAAM;CAI3B"}

package/dist/esm/errors.js

@@ -0,0 +1,62 @@
+/**
+ * Base class for all inference-related errors.
+ */
+export class InferenceClientError extends Error {
+    constructor(message) {
+        super(message);
+        this.name = "InferenceClientError";
+    }
+}
+export class InferenceClientInputError extends InferenceClientError {
+    constructor(message) {
+        super(message);
+        this.name = "InputError";
+    }
+}
+class InferenceClientHttpRequestError extends InferenceClientError {
+    httpRequest;
+    httpResponse;
+    constructor(message, httpRequest, httpResponse) {
+        super(message);
+        this.httpRequest = {
+            ...httpRequest,
+            ...(httpRequest.headers
+                ? {
+                    headers: {
+                        ...httpRequest.headers,
+                        ...("Authorization" in httpRequest.headers ? { Authorization: `Bearer [redacted]` } : undefined),
+                        /// redact authentication in the request headers
+                    },
+                }
+                : undefined),
+        };
+        this.httpResponse = httpResponse;
+    }
+}
+/**
+ * Thrown when the HTTP request to the provider fails, e.g. due to API issues or server errors.
+ */
+export class InferenceClientProviderApiError extends InferenceClientHttpRequestError {
+    constructor(message, httpRequest, httpResponse) {
+        super(message, httpRequest, httpResponse);
+        this.name = "ProviderApiError";
+    }
+}
+/**
+ * Thrown when the HTTP request to the hub fails, e.g. due to API issues or server errors.
+ */
+export class InferenceClientHubApiError extends InferenceClientHttpRequestError {
+    constructor(message, httpRequest, httpResponse) {
+        super(message, httpRequest, httpResponse);
+        this.name = "HubApiError";
+    }
+}
+/**
+ * Thrown when the inference output returned by the provider is invalid / does not match the expectations
+ */
+export class InferenceClientProviderOutputError extends InferenceClientError {
+    constructor(message) {
+        super(message);
+        this.name = "ProviderOutputError";
+    }
+}

package/dist/esm/index.d.ts

@@ -1,5 +1,5 @@
 export { InferenceClient, InferenceClientEndpoint, HfInference } from "./InferenceClient.js";
-export { InferenceOutputError } from "./lib/InferenceOutputError.js";
+export * from "./errors.js";
 export * from "./types.js";
 export * from "./tasks/index.js";
 import * as snippets from "./snippets/index.js";

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

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,uBAAuB,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAC;AAC7F,OAAO,EAAE,oBAAoB,EAAE,MAAM,+BAA+B,CAAC;AACrE,cAAc,YAAY,CAAC;AAC3B,cAAc,kBAAkB,CAAC;AACjC,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAEhD,OAAO,EAAE,QAAQ,EAAE,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,uBAAuB,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAC;AAC7F,cAAc,aAAa,CAAC;AAC5B,cAAc,YAAY,CAAC;AAC3B,cAAc,kBAAkB,CAAC;AACjC,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAEhD,OAAO,EAAE,QAAQ,EAAE,CAAC"}

package/dist/esm/index.js

@@ -1,5 +1,5 @@
 export { InferenceClient, InferenceClientEndpoint, HfInference } from "./InferenceClient.js";
-export { InferenceOutputError } from "./lib/InferenceOutputError.js";
+export * from "./errors.js";
 export * from "./types.js";
 export * from "./tasks/index.js";
 import * as snippets from "./snippets/index.js";

package/dist/esm/lib/getInferenceProviderMapping.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"getInferenceProviderMapping.d.ts","sourceRoot":"","sources":["../../../src/lib/getInferenceProviderMapping.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,oBAAoB,CAAC;AAIrD,OAAO,KAAK,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,OAAO,EAAE,MAAM,aAAa,CAAC;AAGzF,eAAO,MAAM,6BAA6B,uTAA+C,CAAC;AAE1F,MAAM,MAAM,wBAAwB,GAAG,OAAO,CAC7C,MAAM,CAAC,iBAAiB,EAAE,IAAI,CAAC,6BAA6B,EAAE,WAAW,CAAC,CAAC,CAC3E,CAAC;AAEF,MAAM,WAAW,6BAA6B;IAC7C,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,SAAS,EAAE,OAAO,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,IAAI,EAAE,UAAU,CAAC;CACjB;AAED,wBAAsB,qCAAqC,CAC1D,OAAO,EAAE,OAAO,EAChB,WAAW,CAAC,EAAE,MAAM,EACpB,OAAO,CAAC,EAAE;IACT,KAAK,CAAC,EAAE,CAAC,KAAK,EAAE,WAAW,EAAE,IAAI,CAAC,EAAE,WAAW,KAAK,OAAO,CAAC,QAAQ,CAAC,CAAC;CACtE,GACC,OAAO,CAAC,wBAAwB,CAAC,CA6BnC;AAED,wBAAsB,2BAA2B,CAChD,MAAM,EAAE;IACP,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,OAAO,EAAE,OAAO,CAAC;IACjB,QAAQ,EAAE,iBAAiB,CAAC;IAC5B,IAAI,EAAE,UAAU,CAAC;CACjB,EACD,OAAO,EAAE;IACR,KAAK,CAAC,EAAE,CAAC,KAAK,EAAE,WAAW,EAAE,IAAI,CAAC,EAAE,WAAW,KAAK,OAAO,CAAC,QAAQ,CAAC,CAAC;CACtE,GACC,OAAO,CAAC,6BAA6B,GAAG,IAAI,CAAC,CA4B/C;AAED,wBAAsB,eAAe,CACpC,QAAQ,CAAC,EAAE,yBAAyB,EACpC,OAAO,CAAC,EAAE,MAAM,EAChB,WAAW,CAAC,EAAE,MAAM,GAClB,OAAO,CAAC,iBAAiB,CAAC,CAyB5B"}
+{"version":3,"file":"getInferenceProviderMapping.d.ts","sourceRoot":"","sources":["../../../src/lib/getInferenceProviderMapping.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,oBAAoB,CAAC;AAIrD,OAAO,KAAK,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,OAAO,EAAE,MAAM,aAAa,CAAC;AAIzF,eAAO,MAAM,6BAA6B,uTAA+C,CAAC;AAE1F,MAAM,MAAM,wBAAwB,GAAG,OAAO,CAC7C,MAAM,CAAC,iBAAiB,EAAE,IAAI,CAAC,6BAA6B,EAAE,WAAW,CAAC,CAAC,CAC3E,CAAC;AAEF,MAAM,WAAW,6BAA6B;IAC7C,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,SAAS,EAAE,OAAO,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,IAAI,EAAE,UAAU,CAAC;CACjB;AAED,wBAAsB,qCAAqC,CAC1D,OAAO,EAAE,OAAO,EAChB,WAAW,CAAC,EAAE,MAAM,EACpB,OAAO,CAAC,EAAE;IACT,KAAK,CAAC,EAAE,CAAC,KAAK,EAAE,WAAW,EAAE,IAAI,CAAC,EAAE,WAAW,KAAK,OAAO,CAAC,QAAQ,CAAC,CAAC;CACtE,GACC,OAAO,CAAC,wBAAwB,CAAC,CAgDnC;AAED,wBAAsB,2BAA2B,CAChD,MAAM,EAAE;IACP,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,OAAO,EAAE,OAAO,CAAC;IACjB,QAAQ,EAAE,iBAAiB,CAAC;IAC5B,IAAI,EAAE,UAAU,CAAC;CACjB,EACD,OAAO,EAAE;IACR,KAAK,CAAC,EAAE,CAAC,KAAK,EAAE,WAAW,EAAE,IAAI,CAAC,EAAE,WAAW,KAAK,OAAO,CAAC,QAAQ,CAAC,CAAC;CACtE,GACC,OAAO,CAAC,6BAA6B,GAAG,IAAI,CAAC,CA4B/C;AAED,wBAAsB,eAAe,CACpC,QAAQ,CAAC,EAAE,yBAAyB,EACpC,OAAO,CAAC,EAAE,MAAM,EAChB,WAAW,CAAC,EAAE,MAAM,GAClB,OAAO,CAAC,iBAAiB,CAAC,CAyB5B"}

package/dist/esm/lib/getInferenceProviderMapping.js

@@ -2,6 +2,7 @@ import { HF_HUB_URL } from "../config.js";
 import { HARDCODED_MODEL_INFERENCE_MAPPING } from "../providers/consts.js";
 import { EQUIVALENT_SENTENCE_TRANSFORMERS_TASKS } from "../providers/hf-inference.js";
 import { typedInclude } from "../utils/typedInclude.js";
+import { InferenceClientHubApiError, InferenceClientInputError } from "../errors.js";
 export const inferenceProviderMappingCache = new Map();
 export async function fetchInferenceProviderMappingForModel(modelId, accessToken, options) {
     let inferenceProviderMapping;
@@ -10,22 +11,32 @@ export async function fetchInferenceProviderMappingForModel(modelId, accessToken
         inferenceProviderMapping = inferenceProviderMappingCache.get(modelId);
     }
     else {
-        const resp = await (options?.fetch ?? fetch)(`${HF_HUB_URL}/api/models/${modelId}?expand[]=inferenceProviderMapping`, {
+        const url = `${HF_HUB_URL}/api/models/${modelId}?expand[]=inferenceProviderMapping`;
+        const resp = await (options?.fetch ?? fetch)(url, {
             headers: accessToken?.startsWith("hf_") ? { Authorization: `Bearer ${accessToken}` } : {},
         });
-        if (resp.status === 404) {
-            throw new Error(`Model ${modelId} does not exist`);
+        if (!resp.ok) {
+            if (resp.headers.get("Content-Type")?.startsWith("application/json")) {
+                const error = await resp.json();
+                if ("error" in error && typeof error.error === "string") {
+                    throw new InferenceClientHubApiError(`Failed to fetch inference provider mapping for model ${modelId}: ${error.error}`, { url, method: "GET" }, { requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: error });
+                }
+            }
+            else {
+                throw new InferenceClientHubApiError(`Failed to fetch inference provider mapping for model ${modelId}`, { url, method: "GET" }, { requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: await resp.text() });
+            }
         }
-        inferenceProviderMapping = await resp
-            .json()
-            .then((json) => json.inferenceProviderMapping)
-            .catch(() => null);
-        if (inferenceProviderMapping) {
-            inferenceProviderMappingCache.set(modelId, inferenceProviderMapping);
+        let payload = null;
+        try {
+            payload = await resp.json();
         }
-    }
-    if (!inferenceProviderMapping) {
-        throw new Error(`We have not been able to find inference provider information for model ${modelId}.`);
+        catch {
+            throw new InferenceClientHubApiError(`Failed to fetch inference provider mapping for model ${modelId}: malformed API response, invalid JSON`, { url, method: "GET" }, { requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: await resp.text() });
+        }
+        if (!payload?.inferenceProviderMapping) {
+            throw new InferenceClientHubApiError(`We have not been able to find inference provider information for model ${modelId}.`, { url, method: "GET" }, { requestId: resp.headers.get("x-request-id") ?? "", status: resp.status, body: await resp.text() });
+        }
+        inferenceProviderMapping = payload.inferenceProviderMapping;
     }
     return inferenceProviderMapping;
 }
@@ -40,7 +51,7 @@ export async function getInferenceProviderMapping(params, options) {
             ? EQUIVALENT_SENTENCE_TRANSFORMERS_TASKS
             : [params.task];
         if (!typedInclude(equivalentTasks, providerMapping.task)) {
-            throw new Error(`Model ${params.modelId} is not supported for task ${params.task} and provider ${params.provider}. Supported task: ${providerMapping.task}.`);
+            throw new InferenceClientInputError(`Model ${params.modelId} is not supported for task ${params.task} and provider ${params.provider}. Supported task: ${providerMapping.task}.`);
         }
         if (providerMapping.status === "staging") {
             console.warn(`Model ${params.modelId} is in staging mode for provider ${params.provider}. Meant for test purposes only.`);
@@ -52,7 +63,7 @@ export async function getInferenceProviderMapping(params, options) {
 export async function resolveProvider(provider, modelId, endpointUrl) {
     if (endpointUrl) {
         if (provider) {
-            throw new Error("Specifying both endpointUrl and provider is not supported.");
+            throw new InferenceClientInputError("Specifying both endpointUrl and provider is not supported.");
         }
         /// Defaulting to hf-inference helpers / API
         return "hf-inference";
@@ -63,13 +74,13 @@ export async function resolveProvider(provider, modelId, endpointUrl) {
     }
     if (provider === "auto") {
         if (!modelId) {
-            throw new Error("Specifying a model is required when provider is 'auto'");
+            throw new InferenceClientInputError("Specifying a model is required when provider is 'auto'");
         }
         const inferenceProviderMapping = await fetchInferenceProviderMappingForModel(modelId);
         provider = Object.keys(inferenceProviderMapping)[0];
     }
     if (!provider) {
-        throw new Error(`No Inference Provider available for model ${modelId}.`);
+        throw new InferenceClientInputError(`No Inference Provider available for model ${modelId}.`);
     }
     return provider;
 }