feat: add IBM watsonx AI provider support

- Add watsonx to provider names and schemas
- Create watsonx model definitions with IBM Granite models
- Implement WatsonxHandler for API integration
- Update UI components to support watsonx provider
- Add watsonx to MODELS_BY_PROVIDER configuration

Addresses #8087

packages/types/src/provider-settings.ts

@@ -25,6 +25,7 @@ import {
 	vscodeLlmModels,
 	xaiModels,
 	internationalZAiModels,
+	watsonxModels,
 } from "./providers/index.js"
 
 /**
@@ -68,6 +69,7 @@ export const providerNames = [
 	"io-intelligence",
 	"roo",
 	"vercel-ai-gateway",
+	"watsonx",
 ] as const
 
 export const providerNamesSchema = z.enum(providerNames)
@@ -343,6 +345,13 @@ const vercelAiGatewaySchema = baseProviderSettingsSchema.extend({
 	vercelAiGatewayModelId: z.string().optional(),
 })
 
+const watsonxSchema = apiModelIdProviderModelSchema.extend({
+	watsonxApiKey: z.string().optional(),
+	watsonxProjectId: z.string().optional(),
+	watsonxBaseUrl: z.string().optional(),
+	watsonxRegion: z.string().optional(),
+})
+
 const defaultSchema = z.object({
 	apiProvider: z.undefined(),
 })
@@ -384,6 +393,7 @@ export const providerSettingsSchemaDiscriminated = z.discriminatedUnion("apiProv
 	qwenCodeSchema.merge(z.object({ apiProvider: z.literal("qwen-code") })),
 	rooSchema.merge(z.object({ apiProvider: z.literal("roo") })),
 	vercelAiGatewaySchema.merge(z.object({ apiProvider: z.literal("vercel-ai-gateway") })),
+	watsonxSchema.merge(z.object({ apiProvider: z.literal("watsonx") })),
 	defaultSchema,
 ])
 
@@ -425,6 +435,7 @@ export const providerSettingsSchema = z.object({
 	...qwenCodeSchema.shape,
 	...rooSchema.shape,
 	...vercelAiGatewaySchema.shape,
+	...watsonxSchema.shape,
 	...codebaseIndexProviderSchema.shape,
 })
 
@@ -578,6 +589,7 @@ export const MODELS_BY_PROVIDER: Record<
 	unbound: { id: "unbound", label: "Unbound", models: [] },
 	deepinfra: { id: "deepinfra", label: "DeepInfra", models: [] },
 	"vercel-ai-gateway": { id: "vercel-ai-gateway", label: "Vercel AI Gateway", models: [] },
+	watsonx: { id: "watsonx", label: "IBM watsonx", models: Object.keys(watsonxModels) },
 }
 
 export const dynamicProviders = [
@@ -589,6 +601,7 @@ export const dynamicProviders = [
 	"unbound",
 	"deepinfra",
 	"vercel-ai-gateway",
+	"watsonx",
 ] as const satisfies readonly ProviderName[]
 
 export type DynamicProvider = (typeof dynamicProviders)[number]

packages/types/src/providers/index.ts

@@ -30,3 +30,4 @@ export * from "./xai.js"
 export * from "./vercel-ai-gateway.js"
 export * from "./zai.js"
 export * from "./deepinfra.js"
+export * from "./watsonx.js"

packages/types/src/providers/watsonx.ts

@@ -0,0 +1,148 @@
+import type { ModelInfo } from "../model.js"
+
+// IBM watsonx.ai models
+// https://www.ibm.com/products/watsonx-ai
+export type WatsonxModelId = keyof typeof watsonxModels
+
+export const watsonxDefaultModelId: WatsonxModelId = "ibm/granite-3-8b-instruct"
+
+export const watsonxModels = {
+	// Granite models - IBM's foundation models
+	"ibm/granite-3-8b-instruct": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0002,
+		outputPrice: 0.0006,
+		description: "IBM Granite 3.0 8B Instruct - Optimized for enterprise tasks",
+	},
+	"ibm/granite-3-2b-instruct": {
+		maxTokens: 4096,
+		contextWindow: 4096,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0001,
+		outputPrice: 0.0003,
+		description: "IBM Granite 3.0 2B Instruct - Lightweight model for simple tasks",
+	},
+	"ibm/granite-20b-multilingual": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0006,
+		outputPrice: 0.0018,
+		description: "IBM Granite 20B Multilingual - Supports multiple languages",
+	},
+	"ibm/granite-13b-chat-v2": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0004,
+		outputPrice: 0.0012,
+		description: "IBM Granite 13B Chat v2 - Optimized for conversational AI",
+	},
+	"ibm/granite-13b-instruct-v2": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0004,
+		outputPrice: 0.0012,
+		description: "IBM Granite 13B Instruct v2 - General purpose instruction following",
+	},
+	"ibm/granite-7b-lab": {
+		maxTokens: 4096,
+		contextWindow: 4096,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0002,
+		outputPrice: 0.0006,
+		description: "IBM Granite 7B Lab - Experimental model for research",
+	},
+	// Granite Code models - specialized for code generation
+	"ibm/granite-34b-code-instruct": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.001,
+		outputPrice: 0.003,
+		description: "IBM Granite 34B Code Instruct - Specialized for code generation and understanding",
+	},
+	"ibm/granite-20b-code-instruct": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0006,
+		outputPrice: 0.0018,
+		description: "IBM Granite 20B Code Instruct - Code generation model",
+	},
+	"ibm/granite-8b-code-instruct": {
+		maxTokens: 4096,
+		contextWindow: 4096,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0002,
+		outputPrice: 0.0006,
+		description: "IBM Granite 8B Code Instruct - Lightweight code model",
+	},
+	"ibm/granite-3b-code-instruct": {
+		maxTokens: 2048,
+		contextWindow: 2048,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0001,
+		outputPrice: 0.0003,
+		description: "IBM Granite 3B Code Instruct - Fast code completion",
+	},
+	// Third-party models available on watsonx
+	"meta-llama/llama-3-70b-instruct": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0029,
+		outputPrice: 0.0087,
+		description: "Meta Llama 3 70B Instruct on watsonx",
+	},
+	"meta-llama/llama-3-8b-instruct": {
+		maxTokens: 8192,
+		contextWindow: 8192,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0002,
+		outputPrice: 0.0006,
+		description: "Meta Llama 3 8B Instruct on watsonx",
+	},
+	"mistralai/mixtral-8x7b-instruct-v01": {
+		maxTokens: 4096,
+		contextWindow: 32768,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.0005,
+		outputPrice: 0.0015,
+		description: "Mistral Mixtral 8x7B Instruct on watsonx",
+	},
+	"mistralai/mistral-large": {
+		maxTokens: 8192,
+		contextWindow: 32768,
+		supportsImages: false,
+		supportsPromptCache: false,
+		inputPrice: 0.003,
+		outputPrice: 0.009,
+		description: "Mistral Large on watsonx",
+	},
+} as const satisfies Record<string, ModelInfo>
+
+export const watsonxModelInfoSaneDefaults: ModelInfo = {
+	maxTokens: 4096,
+	contextWindow: 8192,
+	supportsImages: false,
+	supportsPromptCache: false,
+	inputPrice: 0,
+	outputPrice: 0,
+}

src/api/index.ts

@@ -40,6 +40,7 @@ import {
 	FeatherlessHandler,
 	VercelAiGatewayHandler,
 	DeepInfraHandler,
+	WatsonxHandler,
 } from "./providers"
 import { NativeOllamaHandler } from "./providers/native-ollama"
 
@@ -165,6 +166,8 @@ export function buildApiHandler(configuration: ProviderSettings): ApiHandler {
 			return new FeatherlessHandler(options)
 		case "vercel-ai-gateway":
 			return new VercelAiGatewayHandler(options)
+		case "watsonx":
+			return new WatsonxHandler(options)
 		default:
 			apiProvider satisfies "gemini-cli" | undefined
 			return new AnthropicHandler(options)

src/api/providers/index.ts

@@ -34,3 +34,4 @@ export { RooHandler } from "./roo"
 export { FeatherlessHandler } from "./featherless"
 export { VercelAiGatewayHandler } from "./vercel-ai-gateway"
 export { DeepInfraHandler } from "./deepinfra"
+export { WatsonxHandler } from "./watsonx"

src/api/providers/watsonx.ts

@@ -0,0 +1,181 @@
+import { Anthropic } from "@anthropic-ai/sdk"
+import OpenAI from "openai"
+
+import {
+	type ModelInfo,
+	type WatsonxModelId,
+	watsonxDefaultModelId,
+	watsonxModels,
+	watsonxModelInfoSaneDefaults,
+} from "@roo-code/types"
+
+import type { ApiHandlerOptions } from "../../shared/api"
+
+import { ApiStream, ApiStreamUsageChunk } from "../transform/stream"
+import { convertToOpenAiMessages } from "../transform/openai-format"
+import { getModelParams } from "../transform/model-params"
+
+import { DEFAULT_HEADERS } from "./constants"
+import { BaseProvider } from "./base-provider"
+import type { SingleCompletionHandler, ApiHandlerCreateMessageMetadata } from "../index"
+import { getApiRequestTimeout } from "./utils/timeout-config"
+import { handleOpenAIError } from "./utils/openai-error-handler"
+
+export class WatsonxHandler extends BaseProvider implements SingleCompletionHandler {
+	protected options: ApiHandlerOptions
+	private client: OpenAI
+	private readonly providerName = "IBM watsonx"
+
+	constructor(options: ApiHandlerOptions) {
+		super()
+		this.options = options
+
+		// Construct the base URL for watsonx API
+		// Default to US South region if not specified
+		const region = this.options.watsonxRegion || "us-south"
+		const baseURL = this.options.watsonxBaseUrl || `https://${region}.ml.cloud.ibm.com/ml/v1`
+		const apiKey = this.options.watsonxApiKey || "not-provided"
+
+		const headers = {
+			...DEFAULT_HEADERS,
+			"X-Watson-Project-Id": this.options.watsonxProjectId || "",
+		}
+
+		const timeout = getApiRequestTimeout()
+
+		this.client = new OpenAI({
+			baseURL,
+			apiKey,
+			defaultHeaders: headers,
+			timeout,
+		})
+	}
+
+	override async *createMessage(
+		systemPrompt: string,
+		messages: Anthropic.Messages.MessageParam[],
+		metadata?: ApiHandlerCreateMessageMetadata,
+	): ApiStream {
+		const { id: modelId, info: modelInfo } = this.getModel()
+
+		// Combine system prompt with messages for watsonx format
+		const systemMessage: OpenAI.Chat.ChatCompletionSystemMessageParam = {
+			role: "system",
+			content: systemPrompt,
+		}
+
+		const convertedMessages = [systemMessage, ...convertToOpenAiMessages(messages)]
+
+		const requestOptions: OpenAI.Chat.Completions.ChatCompletionCreateParamsStreaming = {
+			model: modelId,
+			temperature: this.options.modelTemperature ?? 0.7,
+			messages: convertedMessages,
+			stream: true as const,
+			stream_options: { include_usage: true },
+		}
+
+		// Add max_tokens if needed
+		if (this.options.includeMaxTokens === true) {
+			requestOptions.max_tokens = this.options.modelMaxTokens || modelInfo.maxTokens
+		}
+
+		let stream
+		try {
+			stream = await this.client.chat.completions.create(requestOptions)
+		} catch (error) {
+			throw handleOpenAIError(error, this.providerName)
+		}
+
+		let lastUsage
+
+		for await (const chunk of stream) {
+			const delta = chunk.choices[0]?.delta ?? {}
+
+			if (delta.content) {
+				yield {
+					type: "text",
+					text: delta.content,
+				}
+			}
+
+			if (chunk.usage) {
+				lastUsage = chunk.usage
+			}
+		}
+
+		if (lastUsage) {
+			yield this.processUsageMetrics(lastUsage, modelInfo)
+		}
+	}
+
+	protected processUsageMetrics(usage: any, modelInfo?: ModelInfo): ApiStreamUsageChunk {
+		return {
+			type: "usage",
+			inputTokens: usage?.prompt_tokens || 0,
+			outputTokens: usage?.completion_tokens || 0,
+		}
+	}
+
+	override getModel() {
+		const modelId = this.options.apiModelId
+		let id = modelId && modelId in watsonxModels ? (modelId as WatsonxModelId) : watsonxDefaultModelId
+		let info: ModelInfo = watsonxModels[id]
+		const params = getModelParams({ format: "openai", modelId: id, model: info, settings: this.options })
+		return { id, info, ...params }
+	}
+
+	async completePrompt(prompt: string): Promise<string> {
+		try {
+			const model = this.getModel()
+			const modelInfo = model.info
+
+			const requestOptions: OpenAI.Chat.Completions.ChatCompletionCreateParamsNonStreaming = {
+				model: model.id,
+				messages: [{ role: "user", content: prompt }],
+			}
+
+			// Add max_tokens if needed
+			if (this.options.includeMaxTokens === true) {
+				requestOptions.max_tokens = this.options.modelMaxTokens || modelInfo.maxTokens
+			}
+
+			let response
+			try {
+				response = await this.client.chat.completions.create(requestOptions)
+			} catch (error) {
+				throw handleOpenAIError(error, this.providerName)
+			}
+
+			return response.choices[0]?.message.content || ""
+		} catch (error) {
+			if (error instanceof Error) {
+				throw new Error(`${this.providerName} completion error: ${error.message}`)
+			}
+
+			throw error
+		}
+	}
+}
+
+// Helper function to get available watsonx models (if API endpoint is available)
+export async function getWatsonxModels(
+	baseUrl?: string,
+	apiKey?: string,
+	projectId?: string,
+	region?: string,
+): Promise<string[]> {
+	try {
+		if (!apiKey || !projectId) {
+			return Object.keys(watsonxModels)
+		}
+
+		const actualRegion = region || "us-south"
+		const actualBaseUrl = baseUrl || `https://${actualRegion}.ml.cloud.ibm.com/ml/v1`
+
+		// Note: This is a placeholder - IBM watsonx may have a different endpoint for listing models
+		// For now, return the static list
+		return Object.keys(watsonxModels)
+	} catch (error) {
+		return Object.keys(watsonxModels)
+	}
+}

webview-ui/src/components/ui/hooks/useSelectedModel.ts

@@ -57,6 +57,8 @@ import {
 	vercelAiGatewayDefaultModelId,
 	BEDROCK_CLAUDE_SONNET_4_MODEL_ID,
 	deepInfraDefaultModelId,
+	watsonxDefaultModelId,
+	watsonxModels,
 } from "@roo-code/types"
 
 import type { ModelRecord, RouterModels } from "@roo/api"
@@ -348,11 +350,16 @@ function getSelectedModel({
 			const info = routerModels["vercel-ai-gateway"]?.[id]
 			return { id, info }
 		}
+		case "watsonx": {
+			const id = apiConfiguration.apiModelId ?? watsonxDefaultModelId
+			const info = watsonxModels[id as keyof typeof watsonxModels]
+			return { id, info }
+		}
 		// case "anthropic":
 		// case "human-relay":
 		// case "fake-ai":
 		default: {
-			provider satisfies "anthropic" | "gemini-cli" | "qwen-code" | "human-relay" | "fake-ai"
+			provider satisfies "anthropic" | "gemini-cli" | "human-relay" | "fake-ai"
 			const id = apiConfiguration.apiModelId ?? anthropicDefaultModelId
 			const baseInfo = anthropicModels[id as keyof typeof anthropicModels]