@openclaw/groq-provider 2026.7.2-beta.5 → 2026.7.33

package/dist/api.js

@@ -1,5 +1,5 @@
 //#region extensions/groq/api.ts
-const GROQ_QWEN3_6_27B_ID = "qwen/qwen3.6-27b";
+const GROQ_QWEN3_32B_ID = "qwen/qwen3-32b";
 const GROQ_GPT_OSS_REASONING_IDS = /* @__PURE__ */ new Set([
 	"openai/gpt-oss-20b",
 	"openai/gpt-oss-120b",
@@ -27,7 +27,7 @@ function normalizeGroqModelId(modelId) {
 }
 function resolveGroqReasoningCompatPatch(modelId) {
 	const normalized = normalizeGroqModelId(modelId);
-	if (normalized === GROQ_QWEN3_6_27B_ID) return {
+	if (normalized === GROQ_QWEN3_32B_ID) return {
 		supportsReasoningEffort: true,
 		supportedReasoningEfforts: [...GROQ_QWEN_REASONING_EFFORTS],
 		reasoningEffortMap: GROQ_QWEN_REASONING_EFFORT_MAP

package/dist/index.js

@@ -1,283 +1,38 @@
 import { groqMediaUnderstandingProvider } from "./media-understanding-provider.js";
-import { createAssistantMessageEventStream, streamSimple } from "openclaw/plugin-sdk/llm";
-import { defineSingleProviderPluginEntry } from "openclaw/plugin-sdk/provider-entry";
-//#region extensions/groq/openclaw.plugin.json
-var openclaw_plugin_default = {
-	id: "groq",
-	activation: { "onStartup": false },
-	enabledByDefault: true,
-	providers: ["groq"],
-	providerEndpoints: [{
-		"endpointClass": "groq-native",
-		"hosts": ["api.groq.com"]
-	}],
-	providerRequest: { "providers": { "groq": { "family": "groq" } } },
-	setup: { "providers": [{
-		"id": "groq",
-		"authMethods": ["api-key"],
-		"envVars": ["GROQ_API_KEY"]
-	}] },
-	providerAuthChoices: [{
-		"provider": "groq",
-		"method": "api-key",
-		"choiceId": "groq-api-key",
-		"appGuidedSecret": true,
-		"choiceLabel": "Groq API key",
-		"choiceHint": "Fast OpenAI-compatible inference",
-		"groupId": "groq",
-		"groupLabel": "Groq",
-		"optionKey": "groqApiKey",
-		"cliFlag": "--groq-api-key",
-		"cliOption": "--groq-api-key <key>",
-		"cliDescription": "Groq API key"
-	}],
-	modelCatalog: {
-		"providers": { "groq": {
-			"baseUrl": "https://api.groq.com/openai/v1",
-			"api": "openai-completions",
-			"defaultModel": "openai/gpt-oss-120b",
-			"models": [
-				{
-					"id": "groq/compound",
-					"name": "Compound",
-					"reasoning": false,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 8192,
-					"cost": {
-						"input": 0,
-						"output": 0,
-						"cacheRead": 0,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "groq/compound-mini",
-					"name": "Compound Mini",
-					"reasoning": false,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 8192,
-					"cost": {
-						"input": 0,
-						"output": 0,
-						"cacheRead": 0,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "llama-3.1-8b-instant",
-					"name": "Llama 3.1 8B Instant",
-					"status": "deprecated",
-					"replacedBy": "openai/gpt-oss-20b",
-					"reasoning": false,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 131072,
-					"cost": {
-						"input": .05,
-						"output": .08,
-						"cacheRead": 0,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "llama-3.3-70b-versatile",
-					"name": "Llama 3.3 70B Versatile",
-					"status": "deprecated",
-					"replacedBy": "openai/gpt-oss-120b",
-					"reasoning": false,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 32768,
-					"cost": {
-						"input": .59,
-						"output": .79,
-						"cacheRead": 0,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "openai/gpt-oss-120b",
-					"name": "GPT OSS 120B",
-					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 65536,
-					"cost": {
-						"input": .15,
-						"output": .6,
-						"cacheRead": .075,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "openai/gpt-oss-20b",
-					"name": "GPT OSS 20B",
-					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 65536,
-					"cost": {
-						"input": .075,
-						"output": .3,
-						"cacheRead": .0375,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "openai/gpt-oss-safeguard-20b",
-					"name": "Safety GPT OSS 20B",
-					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 131072,
-					"maxTokens": 65536,
-					"cost": {
-						"input": .075,
-						"output": .3,
-						"cacheRead": .037,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "qwen/qwen3.6-27b",
-					"name": "Qwen 3.6 27B",
-					"reasoning": true,
-					"input": ["text", "image"],
-					"contextWindow": 131072,
-					"maxTokens": 16384,
-					"cost": {
-						"input": .6,
-						"output": 3,
-						"cacheRead": 0,
-						"cacheWrite": 0
-					}
-				}
-			]
-		} },
-		"discovery": { "groq": "refreshable" }
-	},
-	contracts: { "mediaUnderstandingProviders": ["groq"] },
-	mediaUnderstandingProviderMetadata: { "groq": {
-		"capabilities": ["audio"],
-		"defaultModels": { "audio": "whisper-large-v3-turbo" },
-		"autoPriority": { "audio": 20 }
-	} },
-	configSchema: {
-		"type": "object",
-		"additionalProperties": false,
-		"properties": {}
-	}
-};
-//#endregion
+import { definePluginEntry } from "openclaw/plugin-sdk/plugin-entry";
+import { createProviderApiKeyAuthMethod } from "openclaw/plugin-sdk/provider-auth-api-key";
 //#region extensions/groq/index.ts
-const GROQ_OVERSIZED_RECOVERY_MODEL_ID = "llama-3.3-70b-versatile";
-const GROQ_FALLBACK_MAX_TOKENS = 1024;
-function hasWireMaxTokens(value) {
-	if (typeof value !== "object" || value === null || Array.isArray(value)) return false;
-	const record = value;
-	return record.max_completion_tokens !== void 0 || record.max_tokens !== void 0;
-}
-function hasExplicitMaxTokens(extraParams) {
-	return extraParams?.maxTokens !== void 0 || hasWireMaxTokens(extraParams) || hasWireMaxTokens(extraParams?.extra_body) || hasWireMaxTokens(extraParams?.extraBody);
-}
-function isGroqTpmRequestTooLargeEvent(event) {
-	if (event.type !== "error") return false;
-	const message = event.error.errorMessage?.toLowerCase() ?? "";
-	return message.includes("413 request too large for model") && message.includes("tokens per minute (tpm)") && message.includes("limit ") && message.includes("requested ");
-}
-function wrapGroqOversizedRequestRecovery(streamFn, enabled) {
-	const underlying = streamFn ?? streamSimple;
-	if (!enabled) return underlying;
-	const withoutTools = (model, context, options) => {
-		const originalOnPayload = options?.onPayload;
-		return underlying(model, context, {
-			...options,
-			async onPayload(payload, payloadModel) {
-				const replacement = await originalOnPayload?.(payload, payloadModel);
-				const finalPayload = replacement && typeof replacement === "object" ? replacement : payload;
-				const record = finalPayload;
-				delete record.tools;
-				delete record.tool_choice;
-				delete record.parallel_tool_calls;
-				delete record.parallelToolCalls;
-				delete record.max_tokens;
-				delete record.max_completion_tokens;
-				record.max_completion_tokens = GROQ_FALLBACK_MAX_TOKENS;
-				return finalPayload;
-			}
-		});
-	};
-	return (model, context, options) => {
-		if (options?.maxTokens !== void 0) return underlying(model, context, options);
-		const initial = underlying(model, context, options);
-		const output = createAssistantMessageEventStream();
-		const writable = output;
-		(async () => {
-			try {
-				const resolvedInitial = await Promise.resolve(initial);
-				let forwarded = false;
-				let retryWithoutTools = false;
-				for await (const event of resolvedInitial) {
-					if (!forwarded && !options?.signal?.aborted && isGroqTpmRequestTooLargeEvent(event)) {
-						retryWithoutTools = true;
-						break;
-					}
-					writable.push(event);
-					forwarded = true;
-				}
-				if (retryWithoutTools) {
-					const fallback = await Promise.resolve(withoutTools(model, context, options));
-					for await (const event of fallback) writable.push(event);
-				}
-			} catch (error) {
-				writable.push({
-					type: "error",
-					reason: "error",
-					error: {
-						role: "assistant",
-						content: [],
-						api: model.api,
-						provider: model.provider,
-						model: model.id,
-						usage: {
-							input: 0,
-							output: 0,
-							cacheRead: 0,
-							cacheWrite: 0,
-							totalTokens: 0,
-							cost: {
-								input: 0,
-								output: 0,
-								cacheRead: 0,
-								cacheWrite: 0,
-								total: 0
-							}
-						},
-						stopReason: "error",
-						errorMessage: error instanceof Error ? error.message : String(error),
-						timestamp: Date.now()
-					}
-				});
-			} finally {
-				writable.end();
-			}
-		})();
-		return output;
-	};
-}
-var groq_default = defineSingleProviderPluginEntry({
+const GROQ_DEFAULT_MODEL_REF = "groq/llama-3.3-70b-versatile";
+var groq_default = definePluginEntry({
 	id: "groq",
 	name: "Groq Provider",
 	description: "Bundled Groq provider plugin",
-	manifest: openclaw_plugin_default,
-	provider: {
-		label: "Groq",
-		docsPath: "/providers/groq",
-		catalog: { liveModelDiscovery: true },
-		wrapStreamFn: (ctx) => wrapGroqOversizedRequestRecovery(ctx.streamFn, ctx.modelId === GROQ_OVERSIZED_RECOVERY_MODEL_ID && !hasExplicitMaxTokens(ctx.extraParams) && !hasExplicitMaxTokens(ctx.model?.params) && ctx.model?.maxTokensSource === "discovered")
-	},
 	register(api) {
+		api.registerProvider({
+			id: "groq",
+			label: "Groq",
+			docsPath: "/providers/groq",
+			envVars: ["GROQ_API_KEY"],
+			auth: [createProviderApiKeyAuthMethod({
+				providerId: "groq",
+				methodId: "api-key",
+				label: "Groq API key",
+				hint: "Fast OpenAI-compatible inference",
+				optionKey: "groqApiKey",
+				flagName: "--groq-api-key",
+				envVar: "GROQ_API_KEY",
+				promptMessage: "Enter Groq API key",
+				defaultModel: GROQ_DEFAULT_MODEL_REF,
+				wizard: {
+					choiceId: "groq-api-key",
+					choiceLabel: "Groq API key",
+					choiceHint: "Fast OpenAI-compatible inference",
+					groupId: "groq",
+					groupLabel: "Groq",
+					groupHint: "Fast OpenAI-compatible inference"
+				}
+			})]
+		});
 		api.registerMediaUnderstandingProvider(groqMediaUnderstandingProvider);
 	}
 });

package/openclaw.plugin.json

@@ -48,12 +48,11 @@
       "groq": {
         "baseUrl": "https://api.groq.com/openai/v1",
         "api": "openai-completions",
-        "defaultModel": "openai/gpt-oss-120b",
         "models": [
           {
             "id": "groq/compound",
             "name": "Compound",
-            "reasoning": false,
+            "reasoning": true,
             "input": ["text"],
             "contextWindow": 131072,
             "maxTokens": 8192,
@@ -67,7 +66,7 @@
           {
             "id": "groq/compound-mini",
             "name": "Compound Mini",
-            "reasoning": false,
+            "reasoning": true,
             "input": ["text"],
             "contextWindow": 131072,
             "maxTokens": 8192,
@@ -81,8 +80,6 @@
           {
             "id": "llama-3.1-8b-instant",
             "name": "Llama 3.1 8B Instant",
-            "status": "deprecated",
-            "replacedBy": "openai/gpt-oss-20b",
             "reasoning": false,
             "input": ["text"],
             "contextWindow": 131072,
@@ -97,8 +94,6 @@
           {
             "id": "llama-3.3-70b-versatile",
             "name": "Llama 3.3 70B Versatile",
-            "status": "deprecated",
-            "replacedBy": "openai/gpt-oss-120b",
             "reasoning": false,
             "input": ["text"],
             "contextWindow": 131072,
@@ -110,6 +105,23 @@
               "cacheWrite": 0
             }
           },
+          {
+            "id": "meta-llama/llama-4-scout-17b-16e-instruct",
+            "name": "Llama 4 Scout 17B",
+            "reasoning": false,
+            "input": ["text", "image"],
+            "mediaInput": {
+              "image": { "maxPixels": 33177600, "preferredSidePx": 2048, "tokenMode": "provider" }
+            },
+            "contextWindow": 131072,
+            "maxTokens": 8192,
+            "cost": {
+              "input": 0.11,
+              "output": 0.34,
+              "cacheRead": 0,
+              "cacheWrite": 0
+            }
+          },
           {
             "id": "openai/gpt-oss-120b",
             "name": "GPT OSS 120B",
@@ -120,7 +132,7 @@
             "cost": {
               "input": 0.15,
               "output": 0.6,
-              "cacheRead": 0.075,
+              "cacheRead": 0,
               "cacheWrite": 0
             }
           },
@@ -134,7 +146,7 @@
             "cost": {
               "input": 0.075,
               "output": 0.3,
-              "cacheRead": 0.0375,
+              "cacheRead": 0,
               "cacheWrite": 0
             }
           },
@@ -153,15 +165,15 @@
             }
           },
           {
-            "id": "qwen/qwen3.6-27b",
-            "name": "Qwen 3.6 27B",
+            "id": "qwen/qwen3-32b",
+            "name": "Qwen3 32B",
             "reasoning": true,
-            "input": ["text", "image"],
+            "input": ["text"],
             "contextWindow": 131072,
-            "maxTokens": 16384,
+            "maxTokens": 40960,
             "cost": {
-              "input": 0.6,
-              "output": 3,
+              "input": 0.29,
+              "output": 0.59,
               "cacheRead": 0,
               "cacheWrite": 0
             }
@@ -170,7 +182,7 @@
       }
     },
     "discovery": {
-      "groq": "refreshable"
+      "groq": "static"
     }
   },
   "contracts": {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@openclaw/groq-provider",
-  "version": "2026.7.2-beta.5",
+  "version": "2026.7.33",
   "description": "OpenClaw Groq media-understanding provider.",
   "repository": {
     "type": "git",
@@ -18,10 +18,10 @@
       "minHostVersion": ">=2026.6.8"
     },
     "compat": {
-      "pluginApi": ">=2026.7.2-beta.5"
+      "pluginApi": ">=2026.7.33"
     },
     "build": {
-      "openclawVersion": "2026.7.2-beta.5",
+      "openclawVersion": "2026.7.33",
       "bundledDist": false
     },
     "release": {
@@ -38,7 +38,7 @@
     "README.md"
   ],
   "peerDependencies": {
-    "openclaw": ">=2026.7.2-beta.5"
+    "openclaw": ">=2026.7.33"
   },
   "peerDependenciesMeta": {
     "openclaw": {