feat: add GLM-4.5V model support with image capabilities for z.ai provider

packages/types/src/provider-settings.ts

@@ -377,7 +377,7 @@ const sambaNovaSchema = apiModelIdProviderModelSchema.extend({
 	sambaNovaApiKey: z.string().optional(),
 })
 
-export const zaiApiLineSchema = z.enum(["international_coding", "china_coding"])
+export const zaiApiLineSchema = z.enum(["international_coding", "international", "china_coding", "china"])
 
 export type ZaiApiLine = z.infer<typeof zaiApiLineSchema>
 

packages/types/src/providers/zai.ts

@@ -1,8 +1,8 @@
 import type { ModelInfo } from "../model.js"
-import { ZaiApiLine } from "../provider-settings.js"
 
 // Z AI
 // https://docs.z.ai/guides/llm/glm-4.5
+// https://docs.z.ai/guides/llm/glm-4.6
 // https://docs.z.ai/guides/overview/pricing
 
 export type InternationalZAiModelId = keyof typeof internationalZAiModels
@@ -20,6 +20,30 @@ export const internationalZAiModels = {
 		description:
 			"GLM-4.5 is Zhipu's latest featured model. Its comprehensive capabilities in reasoning, coding, and agent reach the state-of-the-art (SOTA) level among open-source models, with a context length of up to 128k.",
 	},
+	"glm-4.6": {
+		maxTokens: 98_304,
+		contextWindow: 204_800,
+		supportsImages: false,
+		supportsPromptCache: true,
+		inputPrice: 0.6,
+		outputPrice: 2.2,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0.11,
+		description:
+			"GLM-4.6 is Zhipu's newest model with an extended context window of up to 200k tokens, providing enhanced capabilities for processing longer documents and conversations.",
+	},
+	"glm-4.5v": {
+		maxTokens: 16_384,
+		contextWindow: 64_000,
+		supportsImages: true,
+		supportsPromptCache: true,
+		inputPrice: 0.6,
+		outputPrice: 1.8,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0.11,
+		// International pricing tiers for GLM-4.5V are not publicly documented as of 2025-10-20; mainland tiers are defined below.
+		description: "GLM-4.5V is Zhipu's new generation of visual reasoning models based on the MOE architecture.",
+	},
 	"glm-4.5-air": {
 		maxTokens: 98_304,
 		contextWindow: 131_072,
@@ -32,17 +56,49 @@ export const internationalZAiModels = {
 		description:
 			"GLM-4.5-Air is the lightweight version of GLM-4.5. It balances performance and cost-effectiveness, and can flexibly switch to hybrid thinking models.",
 	},
-	"glm-4.6": {
+	"glm-4.5-x": {
 		maxTokens: 98_304,
-		contextWindow: 204_800,
+		contextWindow: 131_072,
 		supportsImages: false,
 		supportsPromptCache: true,
-		inputPrice: 0.6,
-		outputPrice: 2.2,
+		inputPrice: 2.2,
+		outputPrice: 8.9,
 		cacheWritesPrice: 0,
-		cacheReadsPrice: 0.11,
-		description:
-			"GLM-4.6 is Zhipu's newest model with an extended context window of up to 200k tokens, providing enhanced capabilities for processing longer documents and conversations.",
+		cacheReadsPrice: 0.45,
+		description: "GLM-4.5-X is the extended version with enhanced capabilities and performance for complex tasks.",
+	},
+	"glm-4.5-airx": {
+		maxTokens: 98_304,
+		contextWindow: 131_072,
+		supportsImages: false,
+		supportsPromptCache: true,
+		inputPrice: 1.1,
+		outputPrice: 4.5,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0.22,
+		description: "GLM-4.5-AirX is the extended version of GLM-4.5-Air with enhanced capabilities.",
+	},
+	"glm-4-32b-0414-128k": {
+		maxTokens: 98_304,
+		contextWindow: 131_072,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.1,
+		outputPrice: 0.1,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0,
+		description: "GLM-4-32B is a 32 billion parameter model with 128k context length, optimized for efficiency.",
+	},
+	"glm-4.5-flash": {
+		maxTokens: 98_304,
+		contextWindow: 131_072,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0,
+		outputPrice: 0,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0,
+		description: "Zhipu's most advanced free model to date.",
 	},
 } as const satisfies Record<string, ModelInfo>
 
@@ -113,6 +169,37 @@ export const mainlandZAiModels = {
 			},
 		],
 	},
+	"glm-4.5v": {
+		maxTokens: 16_384,
+		contextWindow: 64_000,
+		supportsImages: true,
+		supportsPromptCache: true,
+		inputPrice: 0.6,
+		outputPrice: 1.8,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0.11,
+		description: "GLM-4.5V is Zhipu's new generation of visual reasoning models based on the MOE architecture.",
+		tiers: [
+			{
+				contextWindow: 32_000,
+				inputPrice: 0.21,
+				outputPrice: 0.4,
+				cacheReadsPrice: 0.06,
+			},
+			{
+				contextWindow: 64_000,
+				inputPrice: 0.6,
+				outputPrice: 1.8,
+				cacheReadsPrice: 0.11,
+			},
+			{
+				contextWindow: Infinity,
+				inputPrice: 0.6,
+				outputPrice: 1.8,
+				cacheReadsPrice: 0.11,
+			},
+		],
+	},
 	"glm-4.6": {
 		maxTokens: 98_304,
 		contextWindow: 204_800,
@@ -155,15 +242,15 @@ export const mainlandZAiModels = {
 
 export const ZAI_DEFAULT_TEMPERATURE = 0
 
+type ZaiApiLineLiteral = "international_coding" | "international" | "china_coding" | "china"
+
 export const zaiApiLineConfigs = {
 	international_coding: {
 		name: "International Coding Plan",
 		baseUrl: "https://api.z.ai/api/coding/paas/v4",
 		isChina: false,
 	},
-	china_coding: {
-		name: "China Coding Plan",
-		baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-		isChina: true,
-	},
-} satisfies Record<ZaiApiLine, { name: string; baseUrl: string; isChina: boolean }>
+	international: { name: "International Standard", baseUrl: "https://api.z.ai/api/paas/v4", isChina: false },
+	china_coding: { name: "China Coding Plan", baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4", isChina: true },
+	china: { name: "China Standard", baseUrl: "https://open.bigmodel.cn/api/paas/v4", isChina: true },
+} satisfies Record<ZaiApiLineLiteral, { name: string; baseUrl: string; isChina: boolean }>

src/api/providers/__tests__/zai.spec.ts

@@ -84,6 +84,20 @@ describe("ZAiHandler", () => {
 			expect(model.info).toEqual(internationalZAiModels[testModelId])
 			expect(model.info.contextWindow).toBe(204_800)
 		})
+
+		it("should return GLM-4.5V international model with image support", () => {
+			const testModelId: InternationalZAiModelId = "glm-4.5v"
+			const handlerWithModel = new ZAiHandler({
+				apiModelId: testModelId,
+				zaiApiKey: "test-zai-api-key",
+				zaiApiLine: "international",
+			})
+			const model = handlerWithModel.getModel()
+			expect(model.id).toBe(testModelId)
+			expect(model.info).toEqual(internationalZAiModels[testModelId])
+			expect(model.info.supportsImages).toBe(true)
+			expect(model.info.contextWindow).toBe(64_000)
+		})
 	})
 
 	describe("China Z AI", () => {
@@ -134,6 +148,25 @@ describe("ZAiHandler", () => {
 			expect(model.info).toEqual(mainlandZAiModels[testModelId])
 			expect(model.info.contextWindow).toBe(204_800)
 		})
+
+		it("should return GLM-4.5V China model with image support", () => {
+			const testModelId: MainlandZAiModelId = "glm-4.5v"
+			const handlerWithModel = new ZAiHandler({
+				apiModelId: testModelId,
+				zaiApiKey: "test-zai-api-key",
+				zaiApiLine: "china",
+			})
+			const model = handlerWithModel.getModel()
+			expect(model.id).toBe(testModelId)
+			expect(model.info).toEqual(mainlandZAiModels[testModelId])
+			expect(model.info.supportsImages).toBe(true)
+			expect(model.info.contextWindow).toBe(64_000)
+		})
+
+		it("should include tiers for GLM-4.5V China model", () => {
+			const info = mainlandZAiModels["glm-4.5v"]
+			expect(info.tiers?.length).toBe(3)
+		})
 	})
 
 	describe("Default behavior", () => {

src/api/providers/zai.ts

@@ -3,29 +3,47 @@ import {
 	mainlandZAiModels,
 	internationalZAiDefaultModelId,
 	mainlandZAiDefaultModelId,
-	type InternationalZAiModelId,
-	type MainlandZAiModelId,
 	ZAI_DEFAULT_TEMPERATURE,
-	zaiApiLineConfigs,
 } from "@roo-code/types"
 
 import type { ApiHandlerOptions } from "../../shared/api"
+import type { InternationalZAiModelId, MainlandZAiModelId, ModelInfo } from "@roo-code/types"
 
 import { BaseOpenAiCompatibleProvider } from "./base-openai-compatible-provider"
 
-export class ZAiHandler extends BaseOpenAiCompatibleProvider<InternationalZAiModelId | MainlandZAiModelId> {
+type ZAiModelId = InternationalZAiModelId | MainlandZAiModelId
+
+// Local mapping to avoid cross-package runtime dependency issues in CI
+type ZaiApiLineLocal = "international_coding" | "international" | "china_coding" | "china"
+const ZAI_LINE_CONFIGS: Record<ZaiApiLineLocal, { name: string; baseUrl: string; isChina: boolean }> = {
+	international_coding: {
+		name: "International Coding Plan",
+		baseUrl: "https://api.z.ai/api/coding/paas/v4",
+		isChina: false,
+	},
+	international: { name: "International Standard", baseUrl: "https://api.z.ai/api/paas/v4", isChina: false },
+	china_coding: { name: "China Coding Plan", baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4", isChina: true },
+	china: { name: "China Standard", baseUrl: "https://open.bigmodel.cn/api/paas/v4", isChina: true },
+}
+
+export class ZAiHandler extends BaseOpenAiCompatibleProvider<ZAiModelId> {
 	constructor(options: ApiHandlerOptions) {
-		const isChina = zaiApiLineConfigs[options.zaiApiLine ?? "international_coding"].isChina
-		const models = isChina ? mainlandZAiModels : internationalZAiModels
+		const line = (options.zaiApiLine ?? "international_coding") as ZaiApiLineLocal
+		const { isChina, baseUrl } = ZAI_LINE_CONFIGS[line]
+
 		const defaultModelId = isChina ? mainlandZAiDefaultModelId : internationalZAiDefaultModelId
+		const providerModels = (isChina ? mainlandZAiModels : internationalZAiModels) as unknown as Record<
+			ZAiModelId,
+			ModelInfo
+		>
 
 		super({
 			...options,
 			providerName: "Z AI",
-			baseURL: zaiApiLineConfigs[options.zaiApiLine ?? "international_coding"].baseUrl,
+			baseURL: baseUrl,
 			apiKey: options.zaiApiKey ?? "not-provided",
 			defaultProviderModelId: defaultModelId,
-			providerModels: models,
+			providerModels,
 			defaultTemperature: ZAI_DEFAULT_TEMPERATURE,
 		})
 	}