@lobehub/lobehub 2.0.0-next.186 → 2.0.0-next.187

package/packages/model-bank/src/aiModels/aihubmix.ts

@@ -25,7 +25,7 @@ const aihubmixModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-12-11',
     settings: {
-      extendParams: ['gpt5_1ReasoningEffort', 'textVerbosity'],
+      extendParams: ['gpt5_2ReasoningEffort', 'textVerbosity'],
       searchImpl: 'params',
     },
     type: 'chat',
@@ -51,6 +51,7 @@ const aihubmixModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-12-11',
     settings: {
+      extendParams: ['gpt5_2ProReasoningEffort', 'textVerbosity'],
       searchImpl: 'params',
     },
     type: 'chat',
@@ -189,7 +190,8 @@ const aihubmixModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 400_000,
-    description: 'GPT-5 pro uses more compute to think deeper and consistently deliver better answers.',
+    description:
+      'GPT-5 pro uses more compute to think deeper and consistently deliver better answers.',
     displayName: 'GPT-5 pro',
     id: 'gpt-5-pro',
     maxOutput: 272_000,
@@ -478,7 +480,8 @@ const aihubmixModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_047_576,
-    description: 'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
+    description:
+      'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
     displayName: 'GPT-4.1',
     id: 'gpt-4.1',
     maxOutput: 32_768,
@@ -502,8 +505,7 @@ const aihubmixModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_047_576,
-    description:
-      'GPT-4.1 mini balances intelligence, speed, and cost for many use cases.',
+    description: 'GPT-4.1 mini balances intelligence, speed, and cost for many use cases.',
     displayName: 'GPT-4.1 mini',
     id: 'gpt-4.1-mini',
     maxOutput: 32_768,
@@ -1137,6 +1139,41 @@ const aihubmixModels: AIChatModelCard[] = [
       ],
     },
     releasedAt: '2025-11-18',
+    settings: {
+      extendParams: ['thinkingLevel2', 'urlContext'],
+      searchImpl: 'params',
+      searchProvider: 'google',
+    },
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+      search: true,
+      video: true,
+      vision: true,
+    },
+    contextWindowTokens: 1_048_576 + 65_536,
+    description: 'Gemini 3 Flash 是为速度而打造的最智能的模型，将前沿智能与卓越的搜索接地相结合。',
+    displayName: 'Gemini 3 Flash Preview',
+    enabled: true,
+    id: 'gemini-3-flash-preview',
+    maxOutput: 65_536,
+    pricing: {
+      units: [
+        { name: 'textInput_cacheRead', rate: 0.05, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput', rate: 0.5, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 3, strategy: 'fixed', unit: 'millionTokens' },
+        {
+          lookup: { prices: { '1h': 1 }, pricingParams: ['ttl'] },
+          name: 'textInput_cacheWrite',
+          strategy: 'lookup',
+          unit: 'millionTokens',
+        },
+      ],
+    },
+    releasedAt: '2025-12-17',
     settings: {
       extendParams: ['thinkingLevel', 'urlContext'],
       searchImpl: 'params',
@@ -1210,7 +1247,7 @@ const aihubmixModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_048_576 + 65_536,
-    description: "Gemini 2.5 Flash is Google’s best-value model with full capabilities.",
+    description: 'Gemini 2.5 Flash is Google’s best-value model with full capabilities.',
     displayName: 'Gemini 2.5 Flash',
     id: 'gemini-2.5-flash',
     maxOutput: 65_536,
@@ -1238,7 +1275,6 @@ const aihubmixModels: AIChatModelCard[] = [
     description:
       'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, allowing image generation and editing through conversation.',
     displayName: 'Nano Banana',
-    enabled: true,
     id: 'gemini-2.5-flash-image',
     maxOutput: 8192,
     pricing: {
@@ -1262,7 +1298,7 @@ const aihubmixModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Flash-Lite is Google’s smallest, best-value model, designed for large-scale use.",
+      'Gemini 2.5 Flash-Lite is Google’s smallest, best-value model, designed for large-scale use.',
     displayName: 'Gemini 2.5 Flash-Lite',
     id: 'gemini-2.5-flash-lite',
     maxOutput: 65_536,

package/packages/model-bank/src/aiModels/google.ts

@@ -123,7 +123,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 3 Pro is Google’s most powerful agent and vibe-coding model, delivering richer visuals and deeper interaction on top of state-of-the-art reasoning.",
+      'Gemini 3 Pro is Google’s most powerful agent and vibe-coding model, delivering richer visuals and deeper interaction on top of state-of-the-art reasoning.',
     displayName: 'Gemini 3 Pro Preview',
     enabled: true,
     id: 'gemini-3-pro-preview',
@@ -166,6 +166,41 @@ const googleChatModels: AIChatModelCard[] = [
       ],
     },
     releasedAt: '2025-11-18',
+    settings: {
+      extendParams: ['thinkingLevel2', 'urlContext'],
+      searchImpl: 'params',
+      searchProvider: 'google',
+    },
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+      search: true,
+      video: true,
+      vision: true,
+    },
+    contextWindowTokens: 1_048_576 + 65_536,
+    description: 'Gemini 3 Flash 是为速度而打造的最智能的模型，将前沿智能与卓越的搜索接地相结合。',
+    displayName: 'Gemini 3 Flash Preview',
+    enabled: true,
+    id: 'gemini-3-flash-preview',
+    maxOutput: 65_536,
+    pricing: {
+      units: [
+        { name: 'textInput_cacheRead', rate: 0.05, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput', rate: 0.5, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 3, strategy: 'fixed', unit: 'millionTokens' },
+        {
+          lookup: { prices: { '1h': 1 }, pricingParams: ['ttl'] },
+          name: 'textInput_cacheWrite',
+          strategy: 'lookup',
+          unit: 'millionTokens',
+        },
+      ],
+    },
+    releasedAt: '2025-12-17',
     settings: {
       extendParams: ['thinkingLevel', 'urlContext'],
       searchImpl: 'params',
@@ -182,7 +217,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 131_072 + 32_768,
     description:
-      "Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.",
+      'Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.',
     displayName: 'Nano Banana Pro',
     enabled: true,
     id: 'gemini-3-pro-image-preview',
@@ -213,9 +248,8 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Pro is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.",
+      'Gemini 2.5 Pro is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.',
     displayName: 'Gemini 2.5 Pro',
-    enabled: true,
     id: 'gemini-2.5-pro',
     maxOutput: 65_536,
     pricing: {
@@ -273,7 +307,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Pro Preview is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.",
+      'Gemini 2.5 Pro Preview is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.',
     displayName: 'Gemini 2.5 Pro Preview 06-05',
     id: 'gemini-2.5-pro-preview-06-05',
     maxOutput: 65_536,
@@ -326,7 +360,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Pro Preview is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.",
+      'Gemini 2.5 Pro Preview is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.',
     displayName: 'Gemini 2.5 Pro Preview 05-06',
     id: 'gemini-2.5-pro-preview-05-06',
     maxOutput: 65_536,
@@ -377,9 +411,8 @@ const googleChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_048_576 + 65_536,
-    description: "Gemini 2.5 Flash is Google’s best-value model with full capabilities.",
+    description: 'Gemini 2.5 Flash is Google’s best-value model with full capabilities.',
     displayName: 'Gemini 2.5 Flash',
-    enabled: true,
     id: 'gemini-2.5-flash',
     maxOutput: 65_536,
     pricing: {
@@ -432,9 +465,8 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     displayName: 'Nano Banana',
-    enabled: true,
     id: 'gemini-2.5-flash-image',
     maxOutput: 8192,
     pricing: {
@@ -459,7 +491,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     displayName: 'Nano Banana (Preview)',
     id: 'gemini-2.5-flash-image-preview',
     maxOutput: 8192,
@@ -488,7 +520,7 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Flash-Lite is Google’s smallest, best-value model, designed for large-scale use.",
+      'Gemini 2.5 Flash-Lite is Google’s smallest, best-value model, designed for large-scale use.',
     displayName: 'Gemini 2.5 Flash-Lite',
     id: 'gemini-2.5-flash-lite',
     maxOutput: 65_536,
@@ -706,7 +738,8 @@ const googleChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_008_192,
-    description: 'Gemini 1.5 Flash 002 is an efficient multimodal model built for broad deployment.',
+    description:
+      'Gemini 1.5 Flash 002 is an efficient multimodal model built for broad deployment.',
     displayName: 'Gemini 1.5 Flash 002',
     id: 'gemini-1.5-flash-002', // Deprecated on 2025-09-24
     maxOutput: 8192,
@@ -903,7 +936,7 @@ const googleImageModels: AIImageModelCard[] = [
     type: 'image',
     enabled: true,
     description:
-      "Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.",
+      'Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.',
     releasedAt: '2025-11-18',
     parameters: nanoBananaProParameters,
     pricing: {
@@ -918,10 +951,9 @@ const googleImageModels: AIImageModelCard[] = [
   {
     displayName: 'Nano Banana',
     id: 'gemini-2.5-flash-image:image',
-    enabled: true,
     type: 'image',
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     releasedAt: '2025-08-26',
     parameters: nanoBananaParameters,
     pricing: {
@@ -938,7 +970,7 @@ const googleImageModels: AIImageModelCard[] = [
     id: 'gemini-2.5-flash-image-preview:image',
     type: 'image',
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     releasedAt: '2025-08-26',
     parameters: CHAT_MODEL_IMAGE_GENERATION_PARAMS,
     pricing: {

package/packages/model-bank/src/aiModels/hunyuan.ts

@@ -446,6 +446,26 @@ const hunyuanChatModels: AIChatModelCard[] = [
     },
     type: 'chat',
   },
+  {
+    abilities: {
+      vision: true,
+    },
+    contextWindowTokens: 24_000,
+    description:
+      '基于文本 TurboS 基座生产的图生文快思考模型，相比上一版本在图像基础识别、图像分析推理等维度都有明显的效果提升。',
+    displayName: '混元图生文',
+    id: 'hunyuan-vision-1.5-instruct',
+    maxOutput: 16_000,
+    pricing: {
+      currency: 'CNY',
+      units: [
+        { name: 'textInput', rate: 3, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 9, strategy: 'fixed', unit: 'millionTokens' },
+      ],
+    },
+    releasedAt: '2025-12-17',
+    type: 'chat',
+  },
   {
     abilities: {
       vision: true,

package/packages/model-bank/src/aiModels/infiniai.ts

@@ -51,9 +51,29 @@ const infiniaiChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 200_000,
     description:
-      'MiniMax-M2 is a MoE language model optimized for coding and agent workflows, with about 230B total parameters and around 10B active parameters. It strengthens developer scenarios such as multi-file editing, code-run-fix loops, and test verification/repair, while delivering stable, efficient performance in real environments like terminals, IDEs, and CI.',
-    displayName: 'MiniMax M2',
+      'MiniMax-M2.1 是 MiniMax 系列的最新版本，专为多语言编程和真实世界复杂任务优化。作为一款 AI 原生模型，MiniMax-M2.1 在模型性能、智能体框架支持以及多场景适配方面实现了显著提升，旨在帮助企业和个人更快地找到 AI 原生的工作与生活方式。',
+    displayName: 'MiniMax M2.1',
     enabled: true,
+    id: 'minimax-m2.1',
+    maxOutput: 200_000,
+    pricing: {
+      currency: 'CNY',
+      units: [
+        { name: 'textInput', rate: 2.1, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 8.4, strategy: 'fixed', unit: 'millionTokens' },
+      ],
+    },
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 200_000,
+    description:
+      'MiniMax-M2 是一款专为编码与智能体工作流优化的专家混合（MoE）语言模型，具有约 230B 总参数与约 10B 活跃参数。它在保持强通用智能的同时，针对多文件编辑、代码-运行-修复闭环、测试校验修复等开发者场景进行深度增强，在终端、IDE 与 CI 等真实环境中表现稳定、高效。',
+    displayName: 'MiniMax M2',
     id: 'minimax-m2',
     maxOutput: 200_000,
     pricing: {
@@ -71,16 +91,38 @@ const infiniaiChatModels: AIChatModelCard[] = [
       reasoning: true,
     },
     contextWindowTokens: 131_072,
-    description: 'GLM-4.6 is Zhipu AI’s latest LLM, with stronger reasoning and generation.',
-    displayName: 'GLM-4.6',
+    description: 'GLM-4.7 是智谱AI推出的最新大语言模型，具备更强的推理和生成能力。',
+    displayName: 'GLM-4.7',
     enabled: true,
+    id: 'glm-4.7',
+    maxOutput: 4096,
+    pricing: {
+      currency: 'CNY',
+      units: [
+        { name: 'textInput', rate: 3, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 14, strategy: 'fixed', unit: 'millionTokens' },
+      ],
+    },
+    settings: {
+      extendParams: ['enableReasoning'],
+    },
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 131_072,
+    description: 'GLM-4.6 是智谱AI推出的最新大语言模型，具备更强的推理和生成能力。',
+    displayName: 'GLM-4.6',
     id: 'glm-4.6',
     maxOutput: 4096,
     pricing: {
       currency: 'CNY',
       units: [
-        { name: 'textInput', rate: 4, strategy: 'fixed', unit: 'millionTokens' },
-        { name: 'textOutput', rate: 16, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput', rate: 3, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 14, strategy: 'fixed', unit: 'millionTokens' },
       ],
     },
     settings: {
@@ -96,7 +138,6 @@ const infiniaiChatModels: AIChatModelCard[] = [
     contextWindowTokens: 131_072,
     description: 'DeepSeek-V3.2-Exp is an experimental DeepSeek LLM with stronger reasoning and generation.',
     displayName: 'DeepSeek V3.2 Exp',
-    enabled: true,
     id: 'deepseek-v3.2-exp',
     maxOutput: 65_536,
     pricing: {

package/packages/model-bank/src/aiModels/lobehub.ts

@@ -12,7 +12,8 @@ const lobehubChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 400_000,
-    description: "GPT-5.2 — OpenAI's flagship model optimized for coding and agent tasks across industries.",
+    description:
+      "GPT-5.2 — OpenAI's flagship model optimized for coding and agent tasks across industries.",
     displayName: 'GPT-5.2',
     enabled: true,
     id: 'gpt-5.2',
@@ -26,7 +27,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-12-11',
     settings: {
-      extendParams: ['gpt5_1ReasoningEffort', 'textVerbosity'],
+      extendParams: ['gpt5_2ReasoningEffort', 'textVerbosity'],
       searchImpl: 'params',
     },
     type: 'chat',
@@ -164,7 +165,8 @@ const lobehubChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_047_576,
-    description: 'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
+    description:
+      'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
     displayName: 'GPT-4.1',
     enabled: true,
     id: 'gpt-4.1',
@@ -628,7 +630,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 3 Pro is Google’s most powerful agent and vibe-coding model, delivering richer visuals and deeper interaction on top of state-of-the-art reasoning.",
+      'Gemini 3 Pro is Google’s most powerful agent and vibe-coding model, delivering richer visuals and deeper interaction on top of state-of-the-art reasoning.',
     displayName: 'Gemini 3 Pro Preview',
     enabled: true,
     id: 'gemini-3-pro-preview',
@@ -672,7 +674,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-11-18',
     settings: {
-      extendParams: ['thinkingLevel', 'urlContext'],
+      extendParams: ['thinkingLevel2', 'urlContext'],
       searchImpl: 'params',
       searchProvider: 'google',
     },
@@ -728,7 +730,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 1_048_576 + 65_536,
     description:
-      "Gemini 2.5 Pro is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.",
+      'Gemini 2.5 Pro is Google’s most advanced reasoning model, able to reason over code, math, and STEM problems and analyze large datasets, codebases, and documents with long context.',
     displayName: 'Gemini 2.5 Pro',
     enabled: true,
     id: 'gemini-2.5-pro',
@@ -780,7 +782,7 @@ const lobehubChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_048_576 + 65_536,
-    description: "Gemini 2.5 Flash is Google’s best-value model with full capabilities.",
+    description: 'Gemini 2.5 Flash is Google’s best-value model with full capabilities.',
     displayName: 'Gemini 2.5 Flash',
     enabled: true,
     id: 'gemini-2.5-flash',
@@ -819,7 +821,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 131_072 + 32_768,
     description:
-      "Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.",
+      'Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.',
     displayName: 'Nano Banana Pro',
     enabled: true,
     id: 'gemini-3-pro-image-preview',
@@ -845,7 +847,7 @@ const lobehubChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 32_768,
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     displayName: 'Nano Banana',
     enabled: true,
     id: 'gemini-2.5-flash-image-preview',
@@ -1039,7 +1041,7 @@ export const nanoBananaProParameters: ModelParamsSchema = {
 const lobehubImageModels: AIImageModelCard[] = [
   {
     description:
-      "Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.",
+      'Gemini 3 Pro Image (Nano Banana Pro) is Google’s image generation model and also supports multimodal chat.',
     displayName: 'Nano Banana Pro',
     enabled: true,
     id: 'gemini-3-pro-image-preview:image',
@@ -1057,7 +1059,7 @@ const lobehubImageModels: AIImageModelCard[] = [
   },
   {
     description:
-      "Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.",
+      'Nano Banana is Google’s newest, fastest, and most efficient native multimodal model, enabling conversational image generation and editing.',
     displayName: 'Nano Banana',
     enabled: true,
     id: 'gemini-2.5-flash-image-preview:image',

package/packages/model-bank/src/aiModels/minimax.ts

@@ -7,9 +7,53 @@ const minimaxChatModels: AIChatModelCard[] = [
       reasoning: true,
     },
     contextWindowTokens: 204_800,
-    description: 'Built for efficient coding and agent workflows.',
-    displayName: 'MiniMax M2',
+    description: '强大多语言编程实力，全面升级编程体验',
+    displayName: 'MiniMax M2.1',
     enabled: true,
+    id: 'MiniMax-M2.1',
+    maxOutput: 131_072,
+    pricing: {
+      currency: 'CNY',
+      units: [
+        { name: 'textInput_cacheRead', rate: 0.21, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput_cacheWrite', rate: 2.625, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput', rate: 2.1, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 8.4, strategy: 'fixed', unit: 'millionTokens' },
+      ],
+    },
+    releasedAt: '2025-12-23',
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 204_800,
+    description: '强大多语言编程实力，全面升级编程体验。更快，更高效。',
+    displayName: 'MiniMax M2.1 Lightning',
+    id: 'MiniMax-M2.1-Lightning',
+    maxOutput: 131_072,
+    pricing: {
+      currency: 'CNY',
+      units: [
+        { name: 'textInput_cacheRead', rate: 0.21, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput_cacheWrite', rate: 2.625, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textInput', rate: 2.1, strategy: 'fixed', unit: 'millionTokens' },
+        { name: 'textOutput', rate: 16.8, strategy: 'fixed', unit: 'millionTokens' },
+      ],
+    },
+    releasedAt: '2025-12-23',
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 204_800,
+    description: '专为高效编码与Agent工作流而生',
+    displayName: 'MiniMax M2',
     id: 'MiniMax-M2',
     maxOutput: 131_072,
     pricing: {

package/packages/model-bank/src/aiModels/ollamacloud.ts

@@ -42,11 +42,34 @@ const ollamaCloudModels: AIChatModelCard[] = [
     abilities: {
       functionCall: true,
       reasoning: true,
+      vision: true,
     },
-    contextWindowTokens: 200_000,
-    description: 'MiniMax M2 is an efficient LLM built for coding and agent workflows.',
-    displayName: 'MiniMax M2',
+    contextWindowTokens: 1_048_576,
+    description: 'Gemini 3 Flash 是为速度而打造的最智能的模型，将前沿智能与卓越的搜索接地相结合。',
+    displayName: 'Gemini 3 Flash Preview',
+    id: 'gemini-3-flash-preview',
+    releasedAt: '2025-12-17',
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 204_800,
+    displayName: 'MiniMax M2.1',
     enabled: true,
+    id: 'minimax-m2.1',
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 204_800,
+    description: 'MiniMax M2 是专为编码和代理工作流程构建的高效大型语言模型。',
+    displayName: 'MiniMax M2',
     id: 'minimax-m2',
     type: 'chat',
   },
@@ -57,9 +80,21 @@ const ollamaCloudModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 200_000,
     description:
-      'Zhipu’s latest flagship GLM-4.6 (355B) surpasses prior versions in advanced coding, long-text handling, reasoning, and agent capabilities, aligning with Claude Sonnet 4 in programming performance and ranking among the top coding models in China.',
-    displayName: 'GLM-4.6',
+      'GLM-4.7 是智谱最新旗舰模型，面向 Agentic Coding 场景强化了编码能力、长程任务规划与工具协同，并在多个公开基准的当期榜单中取得开源模型中的领先表现。通用能力提升，回复更简洁自然，写作更具沉浸感。在执行复杂智能体任务，在工具调用时指令遵循更强，Artifacts 与 Agentic Coding 的前端美感和长程任务完成效率进一步提升。',
+    displayName: 'GLM-4.7',
     enabled: true,
+    id: 'glm-4.7',
+    type: 'chat',
+  },
+  {
+    abilities: {
+      functionCall: true,
+      reasoning: true,
+    },
+    contextWindowTokens: 200_000,
+    description:
+      '智谱旗舰模型 GLM-4.6 (355B) 在高级编码、长文本处理、推理与智能体能力上全面超越前代，尤其在编程能力上对齐 Claude Sonnet 4，成为国内顶尖的 Coding 模型。',
+    displayName: 'GLM-4.6',
     id: 'glm-4.6',
     type: 'chat',
   },

package/packages/model-bank/src/aiModels/openai.ts

@@ -42,7 +42,7 @@ export const openaiChatModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-12-11',
     settings: {
-      extendParams: ['gpt5_1ReasoningEffort', 'textVerbosity'],
+      extendParams: ['gpt5_2ReasoningEffort', 'textVerbosity'],
       searchImpl: 'params',
     },
     type: 'chat',
@@ -68,6 +68,7 @@ export const openaiChatModels: AIChatModelCard[] = [
     },
     releasedAt: '2025-12-11',
     settings: {
+      extendParams: ['gpt5_2ProReasoningEffort'],
       searchImpl: 'params',
     },
     type: 'chat',
@@ -203,7 +204,8 @@ export const openaiChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 400_000,
-    description: 'GPT-5 pro uses more compute to think deeper and consistently deliver better answers.',
+    description:
+      'GPT-5 pro uses more compute to think deeper and consistently deliver better answers.',
     displayName: 'GPT-5 pro',
     id: 'gpt-5-pro',
     maxOutput: 272_000,
@@ -563,7 +565,8 @@ export const openaiChatModels: AIChatModelCard[] = [
       vision: true,
     },
     contextWindowTokens: 1_047_576,
-    description: 'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
+    description:
+      'GPT-4.1 is our flagship model for complex tasks and cross-domain problem solving.',
     displayName: 'GPT-4.1',
     id: 'gpt-4.1',
     maxOutput: 32_768,

package/packages/model-bank/src/aiModels/qwen.ts

@@ -9,7 +9,7 @@ const qwenChatModels: AIChatModelCard[] = [
       vision: true,
     },
     config: {
-      deploymentName: 'qwen3-vl-plus',
+      deploymentName: 'qwen3-vl-plus-2025-12-19',
     },
     contextWindowTokens: 262_144,
     description: