@openclaw/baseten-provider 2026.9.7 → 2026.10.1-beta.1

package/README.md

@@ -1,6 +1,6 @@
 # OpenClaw Baseten Provider
 
-Official OpenClaw provider plugin for Baseten Model APIs, including Thinking Machines Lab's Inkling.
+Official OpenClaw provider plugin for Baseten Model APIs, with DeepSeek V4.1 Flash as the starter model.
 
 Install from OpenClaw:
 

package/dist/.setup/{models-BmvQ5BVG.mjs → models-BOkRtUeP.mjs}

@@ -1,5 +1,5 @@
 import { buildManifestModelProviderConfig, readManifestProviderDefaultModelRef } from "openclaw/plugin-sdk/provider-catalog-shared";
-import { asNonArrayRecord, asPositiveSafeInteger, filterStringEntries } from "openclaw/plugin-sdk/string-coerce-runtime";
+import { asNonArrayRecord, asOptionalRecord, asPositiveSafeInteger, filterStringEntries, normalizeOptionalString } from "openclaw/plugin-sdk/string-coerce-runtime";
 //#region extensions/baseten/openclaw.plugin.json
 var openclaw_plugin_default = {
 	id: "baseten",
@@ -18,31 +18,27 @@ var openclaw_plugin_default = {
 		"providers": { "baseten": {
 			"baseUrl": "https://inference.baseten.co/v1",
 			"api": "openai-completions",
-			"defaultModel": "thinkingmachines/inkling",
+			"defaultModel": "deepseek-ai/DeepSeek-V4.1-Flash",
 			"models": [
 				{
-					"id": "deepseek-ai/DeepSeek-V4-Pro",
-					"name": "DeepSeek V4 Pro",
+					"id": "deepseek-ai/DeepSeek-V4.1-Flash",
+					"name": "DeepSeek V4.1 Flash",
 					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 262e3,
-					"maxTokens": 262e3,
+					"input": ["text", "image"],
+					"contextWindow": 1048576,
+					"maxTokens": 262144,
 					"cost": {
-						"input": 1.74,
-						"output": 3.48,
-						"cacheRead": .145,
+						"input": .3,
+						"output": 1.2,
+						"cacheRead": .007,
 						"cacheWrite": 0
 					},
 					"compat": {
-						"codeMode": "capable",
 						"supportsReasoningEffort": true,
 						"supportedReasoningEfforts": [
 							"none",
-							"minimal",
 							"low",
-							"medium",
 							"high",
-							"xhigh",
 							"max"
 						],
 						"reasoningEffortMap": {
@@ -53,25 +49,39 @@ var openclaw_plugin_default = {
 					}
 				},
 				{
-					"id": "zai-org/GLM-4.7",
-					"name": "GLM 4.7",
+					"id": "deepseek-ai/DeepSeek-V4-Pro-0813",
+					"name": "DeepSeek V4 Pro 0813",
 					"reasoning": true,
 					"input": ["text"],
-					"contextWindow": 2e5,
-					"maxTokens": 2e5,
+					"contextWindow": 1048576,
+					"maxTokens": 262144,
 					"cost": {
-						"input": .6,
-						"output": 2.2,
-						"cacheRead": .12,
+						"input": 1.32,
+						"output": 3.96,
+						"cacheRead": .132,
 						"cacheWrite": 0
+					},
+					"compat": {
+						"supportsReasoningEffort": true,
+						"supportedReasoningEfforts": [
+							"none",
+							"low",
+							"high",
+							"max"
+						],
+						"reasoningEffortMap": {
+							"off": "none",
+							"none": "none",
+							"adaptive": "max"
+						}
 					}
 				},
 				{
 					"id": "zai-org/GLM-5.2",
 					"name": "GLM 5.2",
 					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 524e3,
+					"input": ["text", "image"],
+					"contextWindow": 1048e3,
 					"maxTokens": 262e3,
 					"cost": {
 						"input": 1.4,
@@ -102,8 +112,8 @@ var openclaw_plugin_default = {
 					"id": "zai-org/GLM-5.2-Fast",
 					"name": "GLM 5.2 Fast",
 					"reasoning": true,
-					"input": ["text"],
-					"contextWindow": 524e3,
+					"input": ["text", "image"],
+					"contextWindow": 1048e3,
 					"maxTokens": 262e3,
 					"cost": {
 						"input": 2.1,
@@ -129,65 +139,6 @@ var openclaw_plugin_default = {
 						}
 					}
 				},
-				{
-					"id": "thinkingmachines/inkling",
-					"name": "Inkling",
-					"reasoning": true,
-					"input": ["text", "image"],
-					"contextWindow": 1048e3,
-					"maxTokens": 32e3,
-					"cost": {
-						"input": 1,
-						"output": 4.05,
-						"cacheRead": .17,
-						"cacheWrite": 0
-					},
-					"compat": {
-						"supportsReasoningEffort": true,
-						"supportedReasoningEfforts": [
-							"none",
-							"minimal",
-							"low",
-							"medium",
-							"high",
-							"xhigh",
-							"max"
-						],
-						"reasoningEffortMap": {
-							"off": "none",
-							"none": "none",
-							"adaptive": "max"
-						}
-					}
-				},
-				{
-					"id": "moonshotai/Kimi-K2.6",
-					"name": "Kimi K2.6",
-					"reasoning": true,
-					"input": ["text", "image"],
-					"contextWindow": 262e3,
-					"maxTokens": 262e3,
-					"cost": {
-						"input": .95,
-						"output": 4,
-						"cacheRead": .16,
-						"cacheWrite": 0
-					}
-				},
-				{
-					"id": "moonshotai/Kimi-K2.7-Code",
-					"name": "Kimi K2.7 Code",
-					"reasoning": true,
-					"input": ["text", "image"],
-					"contextWindow": 262e3,
-					"maxTokens": 262e3,
-					"cost": {
-						"input": .95,
-						"output": 4,
-						"cacheRead": .16,
-						"cacheWrite": 0
-					}
-				},
 				{
 					"id": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B",
 					"name": "Nemotron Ultra",
@@ -212,7 +163,7 @@ var openclaw_plugin_default = {
 					"cost": {
 						"input": .1,
 						"output": .5,
-						"cacheRead": 0,
+						"cacheRead": .1,
 						"cacheWrite": 0
 					},
 					"compat": {
@@ -249,7 +200,7 @@ var openclaw_plugin_default = {
 		"choiceLabel": "Baseten API key",
 		"groupId": "baseten",
 		"groupLabel": "Baseten",
-		"groupHint": "Hosted Model APIs, including Inkling",
+		"groupHint": "Hosted Model APIs",
 		"optionKey": "basetenApiKey",
 		"cliFlag": "--baseten-api-key",
 		"cliOption": "--baseten-api-key <key>",
@@ -263,9 +214,6 @@ var openclaw_plugin_default = {
 };
 //#endregion
 //#region extensions/baseten/models.ts
-/**
-* Baseten model catalog, compat metadata, and live row projection.
-*/
 const BASETEN_MANIFEST_CATALOG = openclaw_plugin_default.modelCatalog.providers.baseten;
 const BASETEN_MODEL_COMPAT = new Map(buildManifestModelProviderConfig({
 	providerId: "baseten",
@@ -289,26 +237,19 @@ const BASE_COMPAT = {
 	supportsTools: true,
 	maxTokensField: "max_tokens"
 };
-/** Base URL for Baseten's OpenAI-compatible Model APIs. */
 const BASETEN_BASE_URL = BASETEN_MANIFEST_CATALOG.baseUrl;
-/** Default Baseten model id used for onboarding. */
 const BASETEN_DEFAULT_MODEL_ID = BASETEN_MANIFEST_CATALOG.defaultModel;
-/** Default Baseten model ref used for onboarding. */
 const BASETEN_DEFAULT_MODEL_REF = readManifestProviderDefaultModelRef(openclaw_plugin_default, "baseten");
-/** Bundled fallback rows for all Baseten Model APIs available at release time. */
 const BASETEN_MODEL_CATALOG = BASETEN_MANIFEST_CATALOG.models;
-/** Whether Baseten requires chat-template thinking control for this model. */
 function usesBasetenChatTemplateThinking(modelId) {
 	return CHAT_TEMPLATE_THINKING_MODEL_IDS.has(modelId.trim().toLowerCase());
 }
-/** Complete OpenAI-compatible transport policy for one Baseten model. */
 function buildBasetenModelCompat(modelId) {
 	return {
 		...BASE_COMPAT,
 		...structuredClone(BASETEN_MODEL_COMPAT.get(modelId))
 	};
 }
-/** Builds the network-free fallback catalog. */
 function buildStaticBasetenModels() {
 	return buildManifestModelProviderConfig({
 		providerId: "baseten",
@@ -333,7 +274,7 @@ function applyLiveReasoningEffortCompat(fallbackCompat, supportsReasoningEffort)
 }
 function projectLiveModel(row, fallback) {
 	if (row.object !== void 0 && row.object !== "model") return;
-	const id = typeof row.id === "string" ? row.id.trim() : "";
+	const id = normalizeOptionalString(row.id);
 	if (!id) return;
 	const hasLiveFeatures = Array.isArray(row.supported_features);
 	const features = new Set(filterStringEntries(row.supported_features));
@@ -341,14 +282,20 @@ function projectLiveModel(row, fallback) {
 	const inputPrice = readPerTokenPrice(pricing.prompt);
 	const outputPrice = readPerTokenPrice(pricing.completion);
 	const cacheReadPrice = readPerTokenPrice(pricing.input_cache_read);
-	const supportsReasoningEffort = features.has("reasoning_effort");
+	const hasDocumentedSparseFeatures = ["deepseek-ai/DeepSeek-V4.1-Flash", "deepseek-ai/DeepSeek-V4-Pro-0813"].includes(id) && [
+		"tools",
+		"reasoning",
+		"json_mode",
+		"structured_outputs"
+	].every((feature) => features.has(feature));
 	const fallbackCompat = fallback?.compat ?? buildBasetenModelCompat(id);
+	const supportsReasoningEffort = features.has("reasoning_effort") || hasDocumentedSparseFeatures && fallbackCompat.supportsReasoningEffort === true;
 	const compat = hasLiveFeatures ? applyLiveReasoningEffortCompat(fallbackCompat, supportsReasoningEffort) : fallbackCompat;
 	return {
 		id,
-		name: typeof row.name === "string" && row.name.trim() ? row.name.trim() : fallback?.name ?? id,
+		name: normalizeOptionalString(row.name) ?? fallback?.name ?? id,
 		reasoning: hasLiveFeatures ? features.has("reasoning") || supportsReasoningEffort : fallback?.reasoning ?? false,
-		input: hasLiveFeatures ? features.has("vision") ? ["text", "image"] : ["text"] : fallback?.input ?? ["text"],
+		input: hasLiveFeatures ? features.has("vision") || hasDocumentedSparseFeatures && fallback?.input.includes("image") ? ["text", "image"] : ["text"] : fallback?.input ?? ["text"],
 		cost: {
 			input: inputPrice ?? fallback?.cost.input ?? 0,
 			output: outputPrice ?? fallback?.cost.output ?? 0,
@@ -365,8 +312,9 @@ function projectBasetenLiveModels(rows) {
 	const fallbacks = new Map(buildStaticBasetenModels().map((model) => [model.id, model]));
 	const seen = /* @__PURE__ */ new Set();
 	const models = [];
-	for (const row of rows) {
-		if (!row || typeof row !== "object" || Array.isArray(row)) continue;
+	for (const value of rows) {
+		const row = asOptionalRecord(value);
+		if (!row) continue;
 		const model = projectLiveModel(row, fallbacks.get(String(row.id)));
 		if (!model || seen.has(model.id)) continue;
 		seen.add(model.id);

package/dist/api.js

@@ -1,4 +1,4 @@
-import { a as buildBasetenModelCompat, c as resolveBasetenDynamicModel, i as BASETEN_MODEL_CATALOG, l as usesBasetenChatTemplateThinking, n as BASETEN_DEFAULT_MODEL_ID, o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, t as BASETEN_BASE_URL } from "./.setup/models-BmvQ5BVG.mjs";
+import { a as buildBasetenModelCompat, c as resolveBasetenDynamicModel, i as BASETEN_MODEL_CATALOG, l as usesBasetenChatTemplateThinking, n as BASETEN_DEFAULT_MODEL_ID, o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, t as BASETEN_BASE_URL } from "./.setup/models-BOkRtUeP.mjs";
 import { applyBasetenConfig } from "./onboard.js";
 import { buildStaticBasetenProvider } from "./provider-catalog.js";
 export { BASETEN_BASE_URL, BASETEN_DEFAULT_MODEL_ID, BASETEN_DEFAULT_MODEL_REF, BASETEN_MODEL_CATALOG, applyBasetenConfig, buildBasetenModelCompat, buildStaticBasetenModels, buildStaticBasetenProvider, resolveBasetenDynamicModel, usesBasetenChatTemplateThinking };

package/dist/index.js

@@ -1,4 +1,4 @@
-import { c as resolveBasetenDynamicModel, s as projectBasetenLiveModels, u as openclaw_plugin_default } from "./.setup/models-BmvQ5BVG.mjs";
+import { c as resolveBasetenDynamicModel, s as projectBasetenLiveModels, u as openclaw_plugin_default } from "./.setup/models-BOkRtUeP.mjs";
 import { applyBasetenSetupConfig } from "./onboard.js";
 import { buildStaticBasetenProvider } from "./provider-catalog.js";
 import { createBasetenThinkingWrapper } from "./stream.js";
@@ -7,7 +7,6 @@ import { buildOpenAICompatibleLiveProviderCatalog } from "openclaw/plugin-sdk/pr
 import { defineSingleProviderPluginEntry } from "openclaw/plugin-sdk/provider-entry";
 import { buildProviderReplayFamilyHooks } from "openclaw/plugin-sdk/provider-model-shared";
 //#region extensions/baseten/index.ts
-/** Baseten provider plugin entrypoint. */
 const PROVIDER_ID = "baseten";
 var baseten_default = defineSingleProviderPluginEntry({
 	id: PROVIDER_ID,
@@ -19,8 +18,9 @@ var baseten_default = defineSingleProviderPluginEntry({
 		docsPath: "/providers/baseten",
 		manifestAuth: {
 			applyConfig: applyBasetenSetupConfig,
+			preserveExistingPrimary: true,
 			noteTitle: "Baseten",
-			noteMessage: ["Baseten hosts Thinking Machines Lab's Inkling and other frontier models behind one OpenAI-compatible API.", "Get your API key at: https://app.baseten.co/settings/api_keys"].join("\n")
+			noteMessage: ["Baseten hosts DeepSeek V4.1 Flash and other models behind one OpenAI-compatible API.", "Get your API key at: https://app.baseten.co/settings/api_keys"].join("\n")
 		},
 		catalog: {
 			order: "simple",

package/dist/models.js

@@ -1,2 +1,2 @@
-import { a as buildBasetenModelCompat, c as resolveBasetenDynamicModel, i as BASETEN_MODEL_CATALOG, l as usesBasetenChatTemplateThinking, n as BASETEN_DEFAULT_MODEL_ID, o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, s as projectBasetenLiveModels, t as BASETEN_BASE_URL } from "./.setup/models-BmvQ5BVG.mjs";
+import { a as buildBasetenModelCompat, c as resolveBasetenDynamicModel, i as BASETEN_MODEL_CATALOG, l as usesBasetenChatTemplateThinking, n as BASETEN_DEFAULT_MODEL_ID, o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, s as projectBasetenLiveModels, t as BASETEN_BASE_URL } from "./.setup/models-BOkRtUeP.mjs";
 export { BASETEN_BASE_URL, BASETEN_DEFAULT_MODEL_ID, BASETEN_DEFAULT_MODEL_REF, BASETEN_MODEL_CATALOG, buildBasetenModelCompat, buildStaticBasetenModels, projectBasetenLiveModels, resolveBasetenDynamicModel, usesBasetenChatTemplateThinking };

package/dist/onboard.js

@@ -1,7 +1,6 @@
-import { o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, t as BASETEN_BASE_URL } from "./.setup/models-BmvQ5BVG.mjs";
+import { o as buildStaticBasetenModels, r as BASETEN_DEFAULT_MODEL_REF, t as BASETEN_BASE_URL } from "./.setup/models-BOkRtUeP.mjs";
 import { createModelCatalogPresetAppliers } from "openclaw/plugin-sdk/provider-onboard";
 //#region extensions/baseten/onboard.ts
-/** Baseten onboarding config helpers. */
 const { applyConfig } = createModelCatalogPresetAppliers({
 	primaryModelRef: BASETEN_DEFAULT_MODEL_REF,
 	resolveParams: (_cfg, catalogModels) => ({
@@ -11,11 +10,11 @@ const { applyConfig } = createModelCatalogPresetAppliers({
 		catalogModels,
 		aliases: [{
 			modelRef: BASETEN_DEFAULT_MODEL_REF,
-			alias: "Inkling"
+			alias: "DeepSeek V4.1 Flash"
 		}]
 	})
 });
-/** Applies Baseten's provider catalog, Inkling alias, and default model. */
+/** Applies Baseten's provider catalog and starter model without replacing an explicit primary. */
 const applyBasetenConfig = (cfg) => applyConfig(cfg, buildStaticBasetenModels());
 const applyBasetenSetupConfig = (cfg) => applyConfig(cfg, cfg.models?.mode === "replace" ? buildStaticBasetenModels() : []);
 //#endregion

package/dist/provider-catalog.js

@@ -1,6 +1,5 @@
-import { o as buildStaticBasetenModels, t as BASETEN_BASE_URL } from "./.setup/models-BmvQ5BVG.mjs";
+import { o as buildStaticBasetenModels, t as BASETEN_BASE_URL } from "./.setup/models-BOkRtUeP.mjs";
 //#region extensions/baseten/provider-catalog.ts
-/** Builds Baseten's network-free fallback provider catalog. */
 function buildStaticBasetenProvider() {
 	return {
 		baseUrl: BASETEN_BASE_URL,

package/dist/stream.js

@@ -1,4 +1,4 @@
-import { l as usesBasetenChatTemplateThinking } from "./.setup/models-BmvQ5BVG.mjs";
+import { l as usesBasetenChatTemplateThinking } from "./.setup/models-BOkRtUeP.mjs";
 import { asNonArrayRecord } from "openclaw/plugin-sdk/string-coerce-runtime";
 import { streamSimple } from "openclaw/plugin-sdk/llm";
 import { normalizeOpenAICompatibleReasoningReplay, streamWithPayloadPatch } from "openclaw/plugin-sdk/provider-stream-shared";
@@ -15,11 +15,15 @@ function createBasetenThinkingWrapper(ctx) {
 			...options,
 			reasoning: thinkingLevel
 		}, (payload) => {
-			if (model.id.trim().toLowerCase() === "deepseek-ai/deepseek-v4-pro") normalizeOpenAICompatibleReasoningReplay(payload, {
-				thinkingEnabled: thinkingLevel !== "off",
-				stripAssistantMessagesOnly: true,
-				replaceNullReasoningContent: true
-			});
+			const normalizedModelId = model.id.trim().toLowerCase();
+			if (normalizedModelId === "deepseek-ai/deepseek-v4-pro" || normalizedModelId === "deepseek-ai/deepseek-v4-pro-0813") {
+				normalizeOpenAICompatibleReasoningReplay(payload, {
+					thinkingEnabled: thinkingLevel !== "off",
+					stripAssistantMessagesOnly: true,
+					replaceNullReasoningContent: true
+				});
+				if (normalizedModelId === "deepseek-ai/deepseek-v4-pro-0813" && payload.reasoning_effort !== void 0) payload.thinking = { type: "enabled" };
+			}
 			if (optIn) payload.chat_template_args = {
 				...asNonArrayRecord(payload.chat_template_args),
 				enable_thinking: thinkingLevel !== "off"

package/dist/thinking.js

@@ -1,4 +1,4 @@
-import { l as usesBasetenChatTemplateThinking } from "./.setup/models-BmvQ5BVG.mjs";
+import { l as usesBasetenChatTemplateThinking } from "./.setup/models-BOkRtUeP.mjs";
 //#region extensions/baseten/thinking.ts
 const BASETEN_BINARY_THINKING_PROFILE = {
 	levels: [{ id: "off" }, {
@@ -15,9 +15,21 @@ const BASETEN_GLM_52_THINKING_PROFILE = {
 	],
 	defaultLevel: "off"
 };
+const BASETEN_DEEPSEEK_THINKING_PROFILE = { levels: [
+	{ id: "off" },
+	{ id: "low" },
+	{ id: "high" },
+	{ id: "max" }
+] };
+const BASETEN_DEEPSEEK_V41_THINKING_PROFILE = {
+	...BASETEN_DEEPSEEK_THINKING_PROFILE,
+	defaultLevel: "high"
+};
 /** Exposes only the thinking levels that Baseten actually accepts for opt-in models. */
 function resolveBasetenThinkingProfile(modelId) {
 	const normalized = modelId.trim().toLowerCase();
+	if (normalized === "deepseek-ai/deepseek-v4.1-flash") return BASETEN_DEEPSEEK_V41_THINKING_PROFILE;
+	if (normalized === "deepseek-ai/deepseek-v4-pro-0813") return BASETEN_DEEPSEEK_THINKING_PROFILE;
 	if (normalized === "zai-org/glm-5.2" || normalized === "zai-org/glm-5.2-fast") return BASETEN_GLM_52_THINKING_PROFILE;
 	return usesBasetenChatTemplateThinking(normalized) ? BASETEN_BINARY_THINKING_PROFILE : void 0;
 }

package/openclaw.plugin.json

@@ -28,44 +28,55 @@
       "baseten": {
         "baseUrl": "https://inference.baseten.co/v1",
         "api": "openai-completions",
-        "defaultModel": "thinkingmachines/inkling",
+        "defaultModel": "deepseek-ai/DeepSeek-V4.1-Flash",
         "models": [
           {
-            "id": "deepseek-ai/DeepSeek-V4-Pro",
-            "name": "DeepSeek V4 Pro",
+            "id": "deepseek-ai/DeepSeek-V4.1-Flash",
+            "name": "DeepSeek V4.1 Flash",
             "reasoning": true,
-            "input": [
-              "text"
-            ],
-            "contextWindow": 262000,
-            "maxTokens": 262000,
+            "input": ["text", "image"],
+            "contextWindow": 1048576,
+            "maxTokens": 262144,
             "cost": {
-              "input": 1.74,
-              "output": 3.48,
-              "cacheRead": 0.145,
+              "input": 0.3,
+              "output": 1.2,
+              "cacheRead": 0.007,
               "cacheWrite": 0
             },
             "compat": {
-              "codeMode": "capable",
               "supportsReasoningEffort": true,
-              "supportedReasoningEfforts": ["none", "minimal", "low", "medium", "high", "xhigh", "max"],
+              "supportedReasoningEfforts": ["none", "low", "high", "max"],
               "reasoningEffortMap": { "off": "none", "none": "none", "adaptive": "max" }
             }
           },
           {
-            "id": "zai-org/GLM-4.7",
-            "name": "GLM 4.7",
+            "id": "deepseek-ai/DeepSeek-V4-Pro-0813",
+            "name": "DeepSeek V4 Pro 0813",
             "reasoning": true,
             "input": [
               "text"
             ],
-            "contextWindow": 200000,
-            "maxTokens": 200000,
+            "contextWindow": 1048576,
+            "maxTokens": 262144,
             "cost": {
-              "input": 0.6,
-              "output": 2.2,
-              "cacheRead": 0.12,
+              "input": 1.32,
+              "output": 3.96,
+              "cacheRead": 0.132,
               "cacheWrite": 0
+            },
+            "compat": {
+              "supportsReasoningEffort": true,
+              "supportedReasoningEfforts": [
+                "none",
+                "low",
+                "high",
+                "max"
+              ],
+              "reasoningEffortMap": {
+                "off": "none",
+                "none": "none",
+                "adaptive": "max"
+              }
             }
           },
           {
@@ -73,9 +84,10 @@
             "name": "GLM 5.2",
             "reasoning": true,
             "input": [
-              "text"
+              "text",
+              "image"
             ],
-            "contextWindow": 524000,
+            "contextWindow": 1048000,
             "maxTokens": 262000,
             "cost": {
               "input": 1.4,
@@ -86,7 +98,11 @@
             "compat": {
               "codeMode": "capable",
               "supportsReasoningEffort": true,
-              "supportedReasoningEfforts": ["none", "high", "max"],
+              "supportedReasoningEfforts": [
+                "none",
+                "high",
+                "max"
+              ],
               "reasoningEffortMap": {
                 "off": "none",
                 "none": "none",
@@ -103,9 +119,10 @@
             "name": "GLM 5.2 Fast",
             "reasoning": true,
             "input": [
-              "text"
+              "text",
+              "image"
             ],
-            "contextWindow": 524000,
+            "contextWindow": 1048000,
             "maxTokens": 262000,
             "cost": {
               "input": 2.1,
@@ -115,7 +132,11 @@
             },
             "compat": {
               "supportsReasoningEffort": true,
-              "supportedReasoningEfforts": ["none", "high", "max"],
+              "supportedReasoningEfforts": [
+                "none",
+                "high",
+                "max"
+              ],
               "reasoningEffortMap": {
                 "off": "none",
                 "none": "none",
@@ -127,62 +148,6 @@
               }
             }
           },
-          {
-            "id": "thinkingmachines/inkling",
-            "name": "Inkling",
-            "reasoning": true,
-            "input": [
-              "text",
-              "image"
-            ],
-            "contextWindow": 1048000,
-            "maxTokens": 32000,
-            "cost": {
-              "input": 1,
-              "output": 4.05,
-              "cacheRead": 0.17,
-              "cacheWrite": 0
-            },
-            "compat": {
-              "supportsReasoningEffort": true,
-              "supportedReasoningEfforts": ["none", "minimal", "low", "medium", "high", "xhigh", "max"],
-              "reasoningEffortMap": { "off": "none", "none": "none", "adaptive": "max" }
-            }
-          },
-          {
-            "id": "moonshotai/Kimi-K2.6",
-            "name": "Kimi K2.6",
-            "reasoning": true,
-            "input": [
-              "text",
-              "image"
-            ],
-            "contextWindow": 262000,
-            "maxTokens": 262000,
-            "cost": {
-              "input": 0.95,
-              "output": 4,
-              "cacheRead": 0.16,
-              "cacheWrite": 0
-            }
-          },
-          {
-            "id": "moonshotai/Kimi-K2.7-Code",
-            "name": "Kimi K2.7 Code",
-            "reasoning": true,
-            "input": [
-              "text",
-              "image"
-            ],
-            "contextWindow": 262000,
-            "maxTokens": 262000,
-            "cost": {
-              "input": 0.95,
-              "output": 4,
-              "cacheRead": 0.16,
-              "cacheWrite": 0
-            }
-          },
           {
             "id": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B",
             "name": "Nemotron Ultra",
@@ -211,13 +176,25 @@
             "cost": {
               "input": 0.1,
               "output": 0.5,
-              "cacheRead": 0,
+              "cacheRead": 0.1,
               "cacheWrite": 0
             },
             "compat": {
               "supportsReasoningEffort": true,
-              "supportedReasoningEfforts": ["none", "minimal", "low", "medium", "high", "xhigh", "max"],
-              "reasoningEffortMap": { "off": "none", "none": "none", "adaptive": "max" }
+              "supportedReasoningEfforts": [
+                "none",
+                "minimal",
+                "low",
+                "medium",
+                "high",
+                "xhigh",
+                "max"
+              ],
+              "reasoningEffortMap": {
+                "off": "none",
+                "none": "none",
+                "adaptive": "max"
+              }
             }
           }
         ]
@@ -246,7 +223,7 @@
       "choiceLabel": "Baseten API key",
       "groupId": "baseten",
       "groupLabel": "Baseten",
-      "groupHint": "Hosted Model APIs, including Inkling",
+      "groupHint": "Hosted Model APIs",
       "optionKey": "basetenApiKey",
       "cliFlag": "--baseten-api-key",
       "cliOption": "--baseten-api-key <key>",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@openclaw/baseten-provider",
-  "version": "2026.9.7",
+  "version": "2026.10.1-beta.1",
   "description": "OpenClaw Baseten provider plugin.",
   "repository": {
     "type": "git",
@@ -18,10 +18,10 @@
       "minHostVersion": ">=2026.7.2"
     },
     "compat": {
-      "pluginApi": ">=2026.9.7"
+      "pluginApi": ">=2026.10.1-beta.1"
     },
     "build": {
-      "openclawVersion": "2026.9.7",
+      "openclawVersion": "2026.10.1-beta.1",
       "bundledDist": false
     },
     "release": {
@@ -40,7 +40,7 @@
     "assets/activity.svg"
   ],
   "peerDependencies": {
-    "openclaw": ">=2026.9.7"
+    "openclaw": ">=2026.10.1-beta.1"
   },
   "peerDependenciesMeta": {
     "openclaw": {