Add Claude Sonnet 4 and Claude Opus 4 models with thinking variants (#3844)

Co-authored-by: cte <[email protected]>

src/api/providers/anthropic.ts

@@ -36,6 +36,8 @@ export class AnthropicHandler extends BaseProvider implements SingleCompletionHa
 		let { id: modelId, maxTokens, thinking, temperature, virtualId } = this.getModel()
 
 		switch (modelId) {
+			case "claude-sonnet-4-20250514":
+			case "claude-opus-4-20250514":
 			case "claude-3-7-sonnet-20250219":
 			case "claude-3-5-sonnet-20241022":
 			case "claude-3-5-haiku-20241022":
@@ -92,13 +94,16 @@ export class AnthropicHandler extends BaseProvider implements SingleCompletionHa
 
 						const betas = []
 
-						// Check for the thinking-128k variant first
+						// Enable extended thinking for Claude 3.7 Sonnet only.
+						// https://docs.anthropic.com/en/docs/about-claude/models/migrating-to-claude-4#extended-output-no-longer-supported
 						if (virtualId === "claude-3-7-sonnet-20250219:thinking") {
 							betas.push("output-128k-2025-02-19")
 						}
 
 						// Then check for models that support prompt caching
 						switch (modelId) {
+							case "claude-sonnet-4-20250514":
+							case "claude-opus-4-20250514":
 							case "claude-3-7-sonnet-20250219":
 							case "claude-3-5-sonnet-20241022":
 							case "claude-3-5-haiku-20241022":
@@ -202,11 +207,14 @@ export class AnthropicHandler extends BaseProvider implements SingleCompletionHa
 		// Track the original model ID for special variant handling
 		const virtualId = id
 
-		// The `:thinking` variant is a virtual identifier for the
-		// `claude-3-7-sonnet-20250219` model with a thinking budget.
+		// The `:thinking` variants are virtual identifiers for models with a thinking budget.
 		// We can handle this more elegantly in the future.
 		if (id === "claude-3-7-sonnet-20250219:thinking") {
 			id = "claude-3-7-sonnet-20250219"
+		} else if (id === "claude-sonnet-4-20250514:thinking") {
+			id = "claude-sonnet-4-20250514"
+		} else if (id === "claude-opus-4-20250514:thinking") {
+			id = "claude-opus-4-20250514"
 		}
 
 		return {

src/shared/api.ts

@@ -9,6 +9,54 @@ export type ApiHandlerOptions = Omit<ProviderSettings, "apiProvider" | "id">
 export type AnthropicModelId = keyof typeof anthropicModels
 export const anthropicDefaultModelId: AnthropicModelId = "claude-3-7-sonnet-20250219"
 export const anthropicModels = {
+	"claude-sonnet-4-20250514:thinking": {
+		maxTokens: 64_000,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 3.0, // $3 per million input tokens
+		outputPrice: 15.0, // $15 per million output tokens
+		cacheWritesPrice: 3.75, // $3.75 per million tokens
+		cacheReadsPrice: 0.3, // $0.30 per million tokens
+		thinking: true,
+	},
+	"claude-sonnet-4-20250514": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 3.0, // $3 per million input tokens
+		outputPrice: 15.0, // $15 per million output tokens
+		cacheWritesPrice: 3.75, // $3.75 per million tokens
+		cacheReadsPrice: 0.3, // $0.30 per million tokens
+		thinking: false,
+	},
+	"claude-opus-4-20250514:thinking": {
+		maxTokens: 64_000,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 15.0, // $15 per million input tokens
+		outputPrice: 75.0, // $75 per million output tokens
+		cacheWritesPrice: 18.75, // $18.75 per million tokens
+		cacheReadsPrice: 1.5, // $1.50 per million tokens
+		thinking: true,
+	},
+	"claude-opus-4-20250514": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 15.0, // $15 per million input tokens
+		outputPrice: 75.0, // $75 per million output tokens
+		cacheWritesPrice: 18.75, // $18.75 per million tokens
+		cacheReadsPrice: 1.5, // $1.50 per million tokens
+		thinking: false,
+	},
 	"claude-3-7-sonnet-20250219:thinking": {
 		maxTokens: 128_000,
 		contextWindow: 200_000,
@@ -160,6 +208,34 @@ export const bedrockModels = {
 		maxCachePoints: 1,
 		cachableFields: ["system"],
 	},
+	"anthropic.claude-sonnet-4-20250514-v1:0": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 3.0,
+		outputPrice: 15.0,
+		cacheWritesPrice: 3.75,
+		cacheReadsPrice: 0.3,
+		minTokensPerCachePoint: 1024,
+		maxCachePoints: 4,
+		cachableFields: ["system", "messages", "tools"],
+	},
+	"anthropic.claude-opus-4-20250514-v1:0": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 15.0,
+		outputPrice: 75.0,
+		cacheWritesPrice: 18.75,
+		cacheReadsPrice: 1.5,
+		minTokensPerCachePoint: 1024,
+		maxCachePoints: 4,
+		cachableFields: ["system", "messages", "tools"],
+	},
 	"anthropic.claude-3-7-sonnet-20250219-v1:0": {
 		maxTokens: 8192,
 		contextWindow: 200_000,
@@ -582,6 +658,54 @@ export const vertexModels = {
 		inputPrice: 1.25,
 		outputPrice: 5,
 	},
+	"claude-sonnet-4-20250514:thinking": {
+		maxTokens: 64_000,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 3.0,
+		outputPrice: 15.0,
+		cacheWritesPrice: 3.75,
+		cacheReadsPrice: 0.3,
+		thinking: true,
+	},
+	"claude-sonnet-4-20250514": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 3.0,
+		outputPrice: 15.0,
+		cacheWritesPrice: 3.75,
+		cacheReadsPrice: 0.3,
+		thinking: false,
+	},
+	"claude-opus-4-20250514:thinking": {
+		maxTokens: 64_000,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 15.0,
+		outputPrice: 75.0,
+		cacheWritesPrice: 18.75,
+		cacheReadsPrice: 1.5,
+		thinking: true,
+	},
+	"claude-opus-4-20250514": {
+		maxTokens: 8192,
+		contextWindow: 200_000,
+		supportsImages: true,
+		supportsComputerUse: true,
+		supportsPromptCache: true,
+		inputPrice: 15.0,
+		outputPrice: 75.0,
+		cacheWritesPrice: 18.75,
+		cacheReadsPrice: 1.5,
+		thinking: false,
+	},
 	"claude-3-7-sonnet@20250219:thinking": {
 		maxTokens: 64_000,
 		contextWindow: 200_000,
@@ -1797,6 +1921,10 @@ export const PROMPT_CACHING_MODELS = new Set([
 	"anthropic/claude-3.7-sonnet",
 	"anthropic/claude-3.7-sonnet:beta",
 	"anthropic/claude-3.7-sonnet:thinking",
+	"anthropic/claude-sonnet-4-20250514",
+	"anthropic/claude-sonnet-4-20250514:thinking",
+	"anthropic/claude-opus-4-20250514",
+	"anthropic/claude-opus-4-20250514:thinking",
 	"google/gemini-2.5-pro-preview",
 	"google/gemini-2.5-flash-preview",
 	"google/gemini-2.5-flash-preview:thinking",
@@ -1814,6 +1942,10 @@ export const COMPUTER_USE_MODELS = new Set([
 	"anthropic/claude-3.7-sonnet",
 	"anthropic/claude-3.7-sonnet:beta",
 	"anthropic/claude-3.7-sonnet:thinking",
+	"anthropic/claude-sonnet-4-20250514",
+	"anthropic/claude-sonnet-4-20250514:thinking",
+	"anthropic/claude-opus-4-20250514",
+	"anthropic/claude-opus-4-20250514:thinking",
 ])
 
 const routerNames = ["openrouter", "requesty", "glama", "unbound", "litellm"] as const