feat: add qwen3-vl-plus vision model support for Qwen Code CLI

- Added qwen3-vl-plus to QwenCodeModelId type
- Added model configuration with vision support (supportsImages: true)
- Set 1M context window and 65K max tokens matching other Qwen models
- Closes #8993

Author: roomote

packages/types/src/providers/qwen-code.ts

@@ -1,6 +1,6 @@
 import type { ModelInfo } from "../model.js"
 
-export type QwenCodeModelId = "qwen3-coder-plus" | "qwen3-coder-flash"
+export type QwenCodeModelId = "qwen3-coder-plus" | "qwen3-coder-flash" | "qwen3-vl-plus"
 
 export const qwenCodeDefaultModelId: QwenCodeModelId = "qwen3-coder-plus"
 
@@ -27,4 +27,15 @@ export const qwenCodeModels = {
 		cacheReadsPrice: 0,
 		description: "Qwen3 Coder Flash - Fast coding model with 1M context window optimized for speed",
 	},
+	"qwen3-vl-plus": {
+		maxTokens: 65_536,
+		contextWindow: 1_000_000,
+		supportsImages: true,
+		supportsPromptCache: false,
+		inputPrice: 0,
+		outputPrice: 0,
+		cacheWritesPrice: 0,
+		cacheReadsPrice: 0,
+		description: "Qwen3 VL Plus - Vision-Language model with 1M context window for multimodal understanding",
+	},
 } as const satisfies Record<QwenCodeModelId, ModelInfo>