feat: 添加Dify平台集成支持

- 升级 dify-sdk-java 从 1.0.7 到 1.2.6 - 新增 ChatModeType.DIFY 枚举类型 - 新增 DifyChatServiceImpl、DifyConversationService、DifyWorkflowService 实现 - 新增 DifyStreamingChatModel 流式聊天模型 - 支持Dify工作流对话模式 - Dify自带RAG知识库时跳过本地向量库查询 Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
fix: 修正application.yml演示模式message缩进
2026-04-24 09:13:39 +00:00 · 2026-04-23 19:19:26 +08:00 · 2026-04-19 13:42:39 +08:00 · 2026-04-19 13:42:05 +08:00 · 2026-04-19 11:47:04 +08:00 · 2026-04-17 18:31:53 +08:00
30 changed files with 1081 additions and 229 deletions
--- a/README.md
+++ b/README.md
@@ -31,7 +31,7 @@

 |     模块     | 现有能力 
 |:----------:|---
-|  **模型管理**  | 多模型接入(OpenAI/DeepSeek/通义/智谱)、多模态理解、Coze/DIFY/FastGPT平台集成 
+|  **模型管理**  | 多模型接入(OpenAI/DeepSeek/通义/智谱/MiniMax)、多模态理解、Coze/DIFY/FastGPT平台集成
 |  **知识管理**  | 本地RAG + 向量库(Milvus/Weaviate/Qdrant)  + 文档解析
 |  **工具管理**  | Mcp协议集成、Skills能力 + 可扩展工具生态                             
 |  **流程编排**  | 可视化工作流设计器、节点拖拽编排、SSE流式执行,目前已经支持模型调用,邮件发送,人工审核等节点  
--- a/README_EN.md
+++ b/README_EN.md
@@ -34,7 +34,7 @@

 | Module | Current Capabilities |
 |:---:|---|
-| **Model Management** | Multi-model integration (OpenAI/DeepSeek/Tongyi/Zhipu), multi-modal understanding, Coze/DIFY/FastGPT platform integration |
+| **Model Management** | Multi-model integration (OpenAI/DeepSeek/Tongyi/Zhipu/MiniMax), multi-modal understanding, Coze/DIFY/FastGPT platform integration |
 | **Knowledge Base** | Local RAG + Vector DB (Milvus/Weaviate/Qdrant) + Document parsing |
 | **Tool Management** | MCP protocol integration, Skills capability + Extensible tool ecosystem |
 | **Workflow Orchestration** | Visual workflow designer, drag-and-drop node orchestration, SSE streaming execution, currently supports model calls, email sending, manual review nodes |
--- a/docs/docker/ruoyi-ai/docker-compose-all.yaml
+++ b/docs/docker/ruoyi-ai/docker-compose-all.yaml
@@ -65,7 +65,7 @@ services:
      - "28080:8080"
    environment:
      QUERY_DEFAULTS_LIMIT: 25
-      AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: true
+      AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: "true"
      PERSISTENCE_DATA_PATH: /var/lib/weaviate
      DEFAULT_VECTORIZER_MODULE: none
      ENABLE_MODULES: text2vec-cohere,text2vec-huggingface,text2vec-palm,text2vec-openai,generative-openai,generative-cohere,generative-palm,ref2vec-centroid,reranker-cohere,qna-openai
--- a/docs/script/sql/ruoyi-ai-v3_mysql8.sql
+++ b/docs/script/sql/ruoyi-ai-v3_mysql8.sql
@@ -72,8 +72,9 @@ CREATE TABLE `chat_model`  (
 -- ----------------------------
 -- Records of chat_model
 -- ----------------------------
-INSERT INTO `chat_model` VALUES (2000585866022060033, 'chat', 'deepseek/deepseek-v3.2', 'ppio', 'deepseek', NULL, 'Y', 'https://api.ppinfra.com/openai', 'sk_xx', 103, 1, '2025-12-15 23:16:54', 1, '2026-03-15 19:18:48', 'DeepSeek-V3.2 是一款在高效推理、复杂推理能力与智能体场景中表现突出的领先模型。其基于 DeepSeek Sparse Attention（DSA）稀疏注意力机制，在显著降低计算开销的同时优化长上下文性能；通过可扩展强化学习框架，整体能力达到 GPT-5 同级，高算力版本 V3.2-Speciale 更在推理表现上接近 Gemini-3.0-Pro；同时，模型依托大型智能体任务合成管线，具备更强的工具调用与多步骤决策能力，并在 2025 年 IMO 与 IOI 中取得金牌级表现。作为 MaaS 平台，我们已对 DeepSeek-V3.2 完成深度适配，通过动态调度、批处理加速、低延迟推理与企业级 SLA 保障，进一步增强其在企业生产环境中的稳定性、性价比与可控性，适用于搜索、问答、智能体、代码、数据处理等多类高价值场景。', 0);
-INSERT INTO `chat_model` VALUES (2007528268536287233, 'vector', 'baai/bge-m3', 'ppio', 'bge-m3', 1024, 'N', 'https://api.ppinfra.com/openai', 'sk_xx', 103, 1, '2026-01-04 03:03:32', 1, '2026-03-15 19:18:51', 'BGE-M3 是一款具备多维度能力的文本嵌入模型，可同时实现密集检索、多向量检索和稀疏检索三大核心功能。该模型设计上兼容超过100种语言，并支持从短句到长达8192词元的长文本等多种输入形式。在跨语言检索任务中，BGE-M3展现出显著优势，其性能在MIRACL、MKQA等国际基准测试中位居前列。此外，针对长文档检索场景，该模型在MLDR、NarritiveQA等数据集上的表现同样达到行业领先水平。', 0);
+INSERT INTO `chat_model` VALUES (2000585866022060033, 'chat', 'zai-org/glm-5', 'ppio', 'zai-org/glm-5', NULL, 'Y', 'https://api.ppio.com/openai', 'sk_xx', 103, 1, '2025-12-15 23:16:54', 1, '2026-03-15 19:18:48', 'DeepSeek-V3.2 是一款在高效推理、复杂推理能力与智能体场景中表现突出的领先模型。其基于 DeepSeek Sparse Attention（DSA）稀疏注意力机制，在显著降低计算开销的同时优化长上下文性能；通过可扩展强化学习框架，整体能力达到 GPT-5 同级，高算力版本 V3.2-Speciale 更在推理表现上接近 Gemini-3.0-Pro；同时，模型依托大型智能体任务合成管线，具备更强的工具调用与多步骤决策能力，并在 2025 年 IMO 与 IOI 中取得金牌级表现。作为 MaaS 平台，我们已对 DeepSeek-V3.2 完成深度适配，通过动态调度、批处理加速、低延迟推理与企业级 SLA 保障，进一步增强其在企业生产环境中的稳定性、性价比与可控性，适用于搜索、问答、智能体、代码、数据处理等多类高价值场景。', 0);
+INSERT INTO `chat_model` VALUES (2007528268536287233, 'vector', 'baai/bge-m3', 'ppio', 'bge-m3', 1024, 'N', 'https://api.ppio.com/openai', 'sk_xx', 103, 1, '2026-01-04 03:03:32', 1, '2026-03-15 19:18:51', 'BGE-M3 是一款具备多维度能力的文本嵌入模型，可同时实现密集检索、多向量检索和稀疏检索三大核心功能。该模型设计上兼容超过100种语言，并支持从短句到长达8192词元的长文本等多种输入形式。在跨语言检索任务中，BGE-M3展现出显著优势，其性能在MIRACL、MKQA等国际基准测试中位居前列。此外，针对长文档检索场景，该模型在MLDR、NarritiveQA等数据集上的表现同样达到行业领先水平。', 0);
+INSERT INTO `chat_model` VALUES (2045735140488847361, 'chat', 'deepseek-chat', 'custom_api', 'deepseek-chat', NULL, NULL, 'https://api.deepseek.com', 'sk_xx', 103, 1, '2026-04-19 13:24:00', 1, '2026-04-19 13:24:00', 'deepseek对话模型', 0);

 -- ----------------------------
 -- Table structure for chat_provider
@@ -95,22 +96,26 @@ CREATE TABLE `chat_provider`  (
                                  `update_time` datetime NULL DEFAULT NULL COMMENT '更新时间',
                                  `remark` varchar(500) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '备注',
                                  `version` int NULL DEFAULT NULL COMMENT '版本',
-                                  `del_flag` char(1) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '0' COMMENT '删除标志（0代表存在 1代表删除）',
+                                  `del_flag` char(1) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '0' COMMENT '删除标志',
                                  `update_ip` varchar(128) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '更新IP',
                                  `tenant_id` bigint NOT NULL DEFAULT 0 COMMENT '租户Id',
                                  PRIMARY KEY (`id`) USING BTREE,
-                                  UNIQUE INDEX `unique_provider_code`(`provider_code` ASC, `tenant_id` ASC) USING BTREE,
+                                  UNIQUE INDEX `unique_provider_code`(`provider_code` ASC, `tenant_id` ASC, `del_flag` ASC) USING BTREE,
                                  INDEX `idx_status`(`status` ASC) USING BTREE
-) ENGINE = InnoDB AUTO_INCREMENT = 2008460994477690882 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_0900_ai_ci COMMENT = '厂商管理表' ROW_FORMAT = DYNAMIC;
+) ENGINE = InnoDB AUTO_INCREMENT = 2045727230803255298 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_0900_ai_ci COMMENT = '厂商管理表' ROW_FORMAT = DYNAMIC;

 -- ----------------------------
 -- Records of chat_provider
 -- ----------------------------
-INSERT INTO `chat_provider` VALUES (1, 'OpenAI', 'openai', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/01091be272334383a1efd9bc22b73ee6.png', 'OpenAI官方API服务商', 'https://api.openai.com', '0', 1, NULL, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:46:59', 'OpenAI厂商', NULL, '0', NULL, 0);
-INSERT INTO `chat_provider` VALUES (2, '阿里云百炼', 'qianwen', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/de2aa7e649de44f3ba5c6380ac6acd04.png', '阿里云百炼大模型服务', 'https://dashscope.aliyuncs.com', '0', 2, NULL, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:49:13', '阿里云厂商', NULL, '0', NULL, 0);
-INSERT INTO `chat_provider` VALUES (3, '智谱AI', 'zhipu', 'https://ruoyi-ai-1254149996.cos.ap-guangzhou.myqcloud.com/2025/12/15/a43e98fb7b3b4861b8caa6184e6fa40a.png', '智谱AI大模型服务', 'https://open.bigmodel.cn', '0', 3, NULL, '2025-12-14 21:48:11', '1', '1', '2026-02-06 00:49:14', '智谱AI厂商', NULL, '1', NULL, 0);
-INSERT INTO `chat_provider` VALUES (5, 'ollama', 'ollama', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/afecabebc8014d80b0f06b4796a74c5d.png', 'ollama大模型', 'http://127.0.0.1:11434', '0', 5, NULL, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:48:48', 'ollama厂商', NULL, '0', NULL, 0);
-INSERT INTO `chat_provider` VALUES (2000585060904435714, 'PPIO', 'ppio', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/049bb6a507174f73bba4b8d8b9e55b8a.png', 'api聚合厂商', 'https://api.ppinfra.com/openai', '0', 5, 103, '2025-12-15 23:13:42', '1', '1', '2026-02-25 20:49:01', 'api聚合厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (1, 'OpenAI', 'openai', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/01091be272334383a1efd9bc22b73ee6.png', 'OpenAI官方API服务商', 'https://api.openai.com', '0', 1, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:46:59', 'OpenAI厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (11, '深度求索', 'deepseek', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/5ba8c30f153246898a4d7dc7b846de8d.png', 'DeepSeek官方API', 'https://api.deepseek.com', '0', 0, 103, '2026-04-19 12:52:34', '1', '1', '2026-04-19 13:13:25', 'DeepSeek官方API', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (12, '智谱AI', 'zhipu', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/da071783c9284fdd9ed1ce1b57b3c75c.png', '智谱AI大模型服务', 'https://open.bigmodel.cn', '0', 4, 103, '2025-12-14 21:48:11', '1', '1', '2026-04-19 13:14:00', '智谱AI厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (13, '小米MIMO', 'xiaomi', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/18dd39365ce244e3ae5e030da036760e.png', '小米官方API', 'https://api.xiaomimimo.com/anthropic/v1/messages', '0', 3, 103, '2026-04-19 12:48:24', '1', '1', '2026-04-19 13:14:22', '小米官方API', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (14, '阿里云百炼', 'qianwen', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/de2aa7e649de44f3ba5c6380ac6acd04.png', '阿里云百炼大模型服务', 'https://dashscope.aliyuncs.com', '0', 2, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:49:13', '阿里云厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (15, 'PPIO', 'ppio', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/049bb6a507174f73bba4b8d8b9e55b8a.png', 'api聚合厂商', 'https://api.ppinfra.com/openai', '0', 5, 103, '2025-12-15 23:13:42', '1', '1', '2026-02-25 20:49:01', 'api聚合厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (16, 'MiniMax', 'minimax', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/fdc712e90e0e4d78b05862ad230884e5.png', 'MiniMax大模型服务，支持M2.7、M2.5等模型', 'https://api.minimax.io/v1', '0', 6, 103, '2026-04-19 12:50:12', '1', '1', '2026-04-19 13:14:59', 'MiniMax厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (17, 'ollama', 'ollama', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/afecabebc8014d80b0f06b4796a74c5d.png', 'ollama大模型', 'http://127.0.0.1:11434', '0', 7, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:48:48', 'ollama厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (18, '自定义厂商', 'custom_api', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/c1a8e122510f4e2f90deb36958af710b.png', 'OPENAI兼容格式', '自定义', '0', 8, 103, '2026-04-19 12:35:57', '1', '1', '2026-04-19 13:17:20', 'OPENAI兼容格式', NULL, '0', NULL, 0);

 -- ----------------------------
 -- Table structure for chat_session
--- a/docs/script/sql/update/updat-0419.sql
+++ b/docs/script/sql/update/updat-0419.sql
@@ -0,0 +1,92 @@
+/*
+ Navicat Premium Dump SQL
+
+ Source Server         : localhost-mysql
+ Source Server Type    : MySQL
+ Source Server Version : 80045 (8.0.45)
+ Source Host           : localhost:3306
+ Source Schema         : ruoyi-ai
+
+ Target Server Type    : MySQL
+ Target Server Version : 80045 (8.0.45)
+ File Encoding         : 65001
+
+ Date: 19/04/2026 13:36:41
+*/
+
+SET NAMES utf8mb4;
+SET FOREIGN_KEY_CHECKS = 0;
+
+-- ----------------------------
+-- Table structure for chat_model
+-- ----------------------------
+DROP TABLE IF EXISTS `chat_model`;
+CREATE TABLE `chat_model`  (
+                               `id` bigint NOT NULL COMMENT '主键',
+                               `category` varchar(20) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '模型分类',
+                               `model_name` varchar(50) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '模型名称',
+                               `provider_code` varchar(20) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '模型供应商',
+                               `model_describe` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '模型描述',
+                               `model_dimension` int NULL DEFAULT NULL COMMENT '模型维度',
+                               `model_show` char(1) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '是否显示',
+                               `api_host` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '请求地址',
+                               `api_key` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '密钥',
+                               `create_dept` bigint NULL DEFAULT NULL COMMENT '创建部门',
+                               `create_by` bigint NULL DEFAULT NULL COMMENT '创建者',
+                               `create_time` datetime NULL DEFAULT NULL COMMENT '创建时间',
+                               `update_by` bigint NULL DEFAULT NULL COMMENT '更新者',
+                               `update_time` datetime NULL DEFAULT NULL COMMENT '更新时间',
+                               `remark` varchar(500) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '备注',
+                               `tenant_id` bigint NOT NULL DEFAULT 0 COMMENT '租户Id',
+                               PRIMARY KEY (`id`) USING BTREE
+) ENGINE = InnoDB CHARACTER SET = utf8mb4 COLLATE = utf8mb4_0900_ai_ci COMMENT = '模型管理' ROW_FORMAT = DYNAMIC;
+
+-- ----------------------------
+-- Records of chat_model
+-- ----------------------------
+INSERT INTO `chat_model` VALUES (2000585866022060033, 'chat', 'zai-org/glm-5', 'ppio', 'zai-org/glm-5', NULL, 'Y', 'https://api.ppio.com/openai', 'sk_xx', 103, 1, '2025-12-15 23:16:54', 1, '2026-03-15 19:18:48', 'DeepSeek-V3.2 是一款在高效推理、复杂推理能力与智能体场景中表现突出的领先模型。其基于 DeepSeek Sparse Attention（DSA）稀疏注意力机制，在显著降低计算开销的同时优化长上下文性能；通过可扩展强化学习框架，整体能力达到 GPT-5 同级，高算力版本 V3.2-Speciale 更在推理表现上接近 Gemini-3.0-Pro；同时，模型依托大型智能体任务合成管线，具备更强的工具调用与多步骤决策能力，并在 2025 年 IMO 与 IOI 中取得金牌级表现。作为 MaaS 平台，我们已对 DeepSeek-V3.2 完成深度适配，通过动态调度、批处理加速、低延迟推理与企业级 SLA 保障，进一步增强其在企业生产环境中的稳定性、性价比与可控性，适用于搜索、问答、智能体、代码、数据处理等多类高价值场景。', 0);
+INSERT INTO `chat_model` VALUES (2007528268536287233, 'vector', 'baai/bge-m3', 'ppio', 'bge-m3', 1024, 'N', 'https://api.ppio.com/openai', 'sk_xx', 103, 1, '2026-01-04 03:03:32', 1, '2026-03-15 19:18:51', 'BGE-M3 是一款具备多维度能力的文本嵌入模型，可同时实现密集检索、多向量检索和稀疏检索三大核心功能。该模型设计上兼容超过100种语言，并支持从短句到长达8192词元的长文本等多种输入形式。在跨语言检索任务中，BGE-M3展现出显著优势，其性能在MIRACL、MKQA等国际基准测试中位居前列。此外，针对长文档检索场景，该模型在MLDR、NarritiveQA等数据集上的表现同样达到行业领先水平。', 0);
+INSERT INTO `chat_model` VALUES (2045735140488847361, 'chat', 'deepseek-chat', 'custom_api', 'deepseek-chat', NULL, NULL, 'https://api.deepseek.com', 'sk_xx', 103, 1, '2026-04-19 13:24:00', 1, '2026-04-19 13:24:00', 'deepseek对话模型', 0);
+
+-- ----------------------------
+-- Table structure for chat_provider
+-- ----------------------------
+DROP TABLE IF EXISTS `chat_provider`;
+CREATE TABLE `chat_provider`  (
+                                  `id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键',
+                                  `provider_name` varchar(100) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NOT NULL COMMENT '厂商名称',
+                                  `provider_code` varchar(50) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NOT NULL COMMENT '厂商编码',
+                                  `provider_icon` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '厂商图标',
+                                  `provider_desc` varchar(500) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '厂商描述',
+                                  `api_host` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT 'API地址',
+                                  `status` char(1) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '0' COMMENT '状态（0正常 1停用）',
+                                  `sort_order` int NULL DEFAULT 0 COMMENT '排序',
+                                  `create_dept` bigint NULL DEFAULT NULL COMMENT '创建部门',
+                                  `create_time` datetime NULL DEFAULT NULL COMMENT '创建时间',
+                                  `create_by` varchar(64) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '' COMMENT '创建者',
+                                  `update_by` varchar(64) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '' COMMENT '更新者',
+                                  `update_time` datetime NULL DEFAULT NULL COMMENT '更新时间',
+                                  `remark` varchar(500) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '备注',
+                                  `version` int NULL DEFAULT NULL COMMENT '版本',
+                                  `del_flag` char(1) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT '0' COMMENT '删除标志',
+                                  `update_ip` varchar(128) CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci NULL DEFAULT NULL COMMENT '更新IP',
+                                  `tenant_id` bigint NOT NULL DEFAULT 0 COMMENT '租户Id',
+                                  PRIMARY KEY (`id`) USING BTREE,
+                                  UNIQUE INDEX `unique_provider_code`(`provider_code` ASC, `tenant_id` ASC, `del_flag` ASC) USING BTREE,
+                                  INDEX `idx_status`(`status` ASC) USING BTREE
+) ENGINE = InnoDB AUTO_INCREMENT = 2045727230803255298 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_0900_ai_ci COMMENT = '厂商管理表' ROW_FORMAT = DYNAMIC;
+
+-- ----------------------------
+-- Records of chat_provider
+-- ----------------------------
+INSERT INTO `chat_provider` VALUES (1, 'OpenAI', 'openai', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/01091be272334383a1efd9bc22b73ee6.png', 'OpenAI官方API服务商', 'https://api.openai.com', '0', 1, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:46:59', 'OpenAI厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (11, '深度求索', 'deepseek', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/5ba8c30f153246898a4d7dc7b846de8d.png', 'DeepSeek官方API', 'https://api.deepseek.com', '0', 0, 103, '2026-04-19 12:52:34', '1', '1', '2026-04-19 13:13:25', 'DeepSeek官方API', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (12, '智谱AI', 'zhipu', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/da071783c9284fdd9ed1ce1b57b3c75c.png', '智谱AI大模型服务', 'https://open.bigmodel.cn', '0', 4, 103, '2025-12-14 21:48:11', '1', '1', '2026-04-19 13:14:00', '智谱AI厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (13, '小米MIMO', 'xiaomi', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/18dd39365ce244e3ae5e030da036760e.png', '小米官方API', 'https://api.xiaomimimo.com/anthropic/v1/messages', '0', 3, 103, '2026-04-19 12:48:24', '1', '1', '2026-04-19 13:14:22', '小米官方API', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (14, '阿里云百炼', 'qianwen', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/de2aa7e649de44f3ba5c6380ac6acd04.png', '阿里云百炼大模型服务', 'https://dashscope.aliyuncs.com', '0', 2, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:49:13', '阿里云厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (15, 'PPIO', 'ppio', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/049bb6a507174f73bba4b8d8b9e55b8a.png', 'api聚合厂商', 'https://api.ppinfra.com/openai', '0', 5, 103, '2025-12-15 23:13:42', '1', '1', '2026-02-25 20:49:01', 'api聚合厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (16, 'MiniMax', 'minimax', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/fdc712e90e0e4d78b05862ad230884e5.png', 'MiniMax大模型服务，支持M2.7、M2.5等模型', 'https://api.minimax.io/v1', '0', 6, 103, '2026-04-19 12:50:12', '1', '1', '2026-04-19 13:14:59', 'MiniMax厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (17, 'ollama', 'ollama', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/02/25/afecabebc8014d80b0f06b4796a74c5d.png', 'ollama大模型', 'http://127.0.0.1:11434', '0', 7, 103, '2025-12-14 21:48:11', '1', '1', '2026-02-25 20:48:48', 'ollama厂商', NULL, '0', NULL, 0);
+INSERT INTO `chat_provider` VALUES (18, '自定义厂商', 'custom_api', 'https://ruoyiai-1254149996.cos.ap-guangzhou.myqcloud.com/2026/04/19/c1a8e122510f4e2f90deb36958af710b.png', 'OPENAI兼容格式', '自定义', '0', 8, 103, '2026-04-19 12:35:57', '1', '1', '2026-04-19 13:17:20', 'OPENAI兼容格式', NULL, '0', NULL, 0);
+
+SET FOREIGN_KEY_CHECKS = 1;
--- a/pom.xml
+++ b/pom.xml
@@ -58,7 +58,7 @@
        <langchain4j.community.version>1.13.0-beta23</langchain4j.community.version>
        <langgraph4j.version>1.5.3</langgraph4j.version>
        <weaviate.version>1.19.6</weaviate.version>
-        <dify.version>1.0.7</dify.version>
+        <dify.version>1.2.6</dify.version>
        <!-- gRPC 版本 - 解决 Milvus SDK 依赖冲突 -->
        <grpc.version>1.62.2</grpc.version>
        <!-- Apache Commons Compress - 用于POI处理ZIP格式 -->
--- a/ruoyi-admin/src/main/resources/application.yml
+++ b/ruoyi-admin/src/main/resources/application.yml
@@ -265,7 +265,7 @@ demo:
  # 是否开启演示模式（开启后所有写操作将被拦截）
  enabled: false
  # 提示消息
-  message: "演示模式，不允许进行写操作"
+  message: "演示模式，不允许操作"
  # 排除的路径（这些路径不受演示模式限制）
  excludes:
    - /login
@@ -276,7 +276,9 @@ demo:
    - /chat/send
    - /system/session/**
    - /system/message/**
-
+    - /system/attach/**
+    - /system/fragment/**
+    - /system/info/**
 --- # warm-flow工作流配置
 warm-flow:
  # 是否开启工作流，默认true
--- a/ruoyi-modules/ruoyi-chat/pom.xml
+++ b/ruoyi-modules/ruoyi-chat/pom.xml
@@ -174,6 +174,23 @@
            <scope>test</scope>
        </dependency>

+        <!-- Test dependencies -->
+        <dependency>
+            <groupId>org.junit.jupiter</groupId>
+            <artifactId>junit-jupiter</artifactId>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>org.mockito</groupId>
+            <artifactId>mockito-core</artifactId>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>org.mockito</groupId>
+            <artifactId>mockito-junit-jupiter</artifactId>
+            <scope>test</scope>
+        </dependency>
+
    </dependencies>

 </project>
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/chat/ChatModelController.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/chat/ChatModelController.java
@@ -9,6 +9,7 @@ import cn.dev33.satoken.annotation.SaCheckPermission;
 import org.ruoyi.common.chat.service.chat.IChatModelService;
 import org.ruoyi.common.chat.domain.bo.chat.ChatModelBo;
 import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
 import org.ruoyi.enums.ModelType;
 import org.springframework.web.bind.annotation.*;
 import org.springframework.validation.annotation.Validated;
@@ -23,6 +24,8 @@ import org.ruoyi.common.log.enums.BusinessType;
 import org.ruoyi.common.excel.utils.ExcelUtil;
 import org.ruoyi.common.mybatis.core.page.TableDataInfo;

+import java.util.LinkedHashMap;
+
 /**
 * 模型管理
 *
@@ -55,6 +58,21 @@ public class ChatModelController extends BaseController {
        return R.ok(chatModelService.queryList(bo));
    }

+    /**
+     * 获取模型供应商枚举
+     */
+    @GetMapping("/providerOptions")
+    public R<List<LinkedHashMap<String, String>>> providerOptions() {
+        List<LinkedHashMap<String, String>> options = new java.util.ArrayList<>();
+        for (ChatModeType type : ChatModeType.values()) {
+            LinkedHashMap<String, String> item = new LinkedHashMap<>();
+            item.put("label", type.getDescription());
+            item.put("value", type.getCode());
+            options.add(item);
+        }
+        return R.ok(options);
+    }
+
    /**
     * 导出模型管理列表
     */
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/knowledge/KnowledgeGraphInstanceController.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/knowledge/KnowledgeGraphInstanceController.java
@@ -1,105 +0,0 @@
-package org.ruoyi.controller.knowledge;
-
-import java.util.List;
-
-import lombok.RequiredArgsConstructor;
-import jakarta.servlet.http.HttpServletResponse;
-import jakarta.validation.constraints.*;
-import cn.dev33.satoken.annotation.SaCheckPermission;
-import org.ruoyi.domain.bo.knowledge.KnowledgeGraphInstanceBo;
-import org.ruoyi.domain.vo.knowledge.KnowledgeGraphInstanceVo;
-import org.ruoyi.service.knowledge.IKnowledgeGraphInstanceService;
-import org.springframework.web.bind.annotation.*;
-import org.springframework.validation.annotation.Validated;
-import org.ruoyi.common.idempotent.annotation.RepeatSubmit;
-import org.ruoyi.common.log.annotation.Log;
-import org.ruoyi.common.web.core.BaseController;
-import org.ruoyi.common.mybatis.core.page.PageQuery;
-import org.ruoyi.common.core.domain.R;
-import org.ruoyi.common.core.validate.AddGroup;
-import org.ruoyi.common.core.validate.EditGroup;
-import org.ruoyi.common.log.enums.BusinessType;
-import org.ruoyi.common.excel.utils.ExcelUtil;
-import org.ruoyi.common.mybatis.core.page.TableDataInfo;
-
-/**
- * 知识图谱实例
- *
- * @author ageerle
- * @date 2025-12-17
- */
-@Validated
-@RequiredArgsConstructor
-@RestController
-@RequestMapping("/system/graphInstance")
-public class KnowledgeGraphInstanceController extends BaseController {
-
-    private final IKnowledgeGraphInstanceService knowledgeGraphInstanceService;
-
-    /**
-     * 查询知识图谱实例列表
-     */
-    @SaCheckPermission("system:graphInstance:list")
-    @GetMapping("/list")
-    public TableDataInfo<KnowledgeGraphInstanceVo> list(KnowledgeGraphInstanceBo bo, PageQuery pageQuery) {
-        return knowledgeGraphInstanceService.queryPageList(bo, pageQuery);
-    }
-
-    /**
-     * 导出知识图谱实例列表
-     */
-    @SaCheckPermission("system:graphInstance:export")
-    @Log(title = "知识图谱实例", businessType = BusinessType.EXPORT)
-    @PostMapping("/export")
-    public void export(KnowledgeGraphInstanceBo bo, HttpServletResponse response) {
-        List<KnowledgeGraphInstanceVo> list = knowledgeGraphInstanceService.queryList(bo);
-        ExcelUtil.exportExcel(list, "知识图谱实例", KnowledgeGraphInstanceVo.class, response);
-    }
-
-    /**
-     * 获取知识图谱实例详细信息
-     *
-     * @param id 主键
-     */
-    @SaCheckPermission("system:graphInstance:query")
-    @GetMapping("/{id}")
-    public R<KnowledgeGraphInstanceVo> getInfo(@NotNull(message = "主键不能为空")
-                                     @PathVariable Long id) {
-        return R.ok(knowledgeGraphInstanceService.queryById(id));
-    }
-
-    /**
-     * 新增知识图谱实例
-     */
-    @SaCheckPermission("system:graphInstance:add")
-    @Log(title = "知识图谱实例", businessType = BusinessType.INSERT)
-    @RepeatSubmit()
-    @PostMapping()
-    public R<Void> add(@Validated(AddGroup.class) @RequestBody KnowledgeGraphInstanceBo bo) {
-        return toAjax(knowledgeGraphInstanceService.insertByBo(bo));
-    }
-
-    /**
-     * 修改知识图谱实例
-     */
-    @SaCheckPermission("system:graphInstance:edit")
-    @Log(title = "知识图谱实例", businessType = BusinessType.UPDATE)
-    @RepeatSubmit()
-    @PutMapping()
-    public R<Void> edit(@Validated(EditGroup.class) @RequestBody KnowledgeGraphInstanceBo bo) {
-        return toAjax(knowledgeGraphInstanceService.updateByBo(bo));
-    }
-
-    /**
-     * 删除知识图谱实例
-     *
-     * @param ids 主键串
-     */
-    @SaCheckPermission("system:graphInstance:remove")
-    @Log(title = "知识图谱实例", businessType = BusinessType.DELETE)
-    @DeleteMapping("/{ids}")
-    public R<Void> remove(@NotEmpty(message = "主键不能为空")
-                          @PathVariable Long[] ids) {
-        return toAjax(knowledgeGraphInstanceService.deleteWithValidByIds(List.of(ids), true));
-    }
-}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/knowledge/KnowledgeGraphSegmentController.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/controller/knowledge/KnowledgeGraphSegmentController.java
@@ -1,105 +0,0 @@
-package org.ruoyi.controller.knowledge;
-
-import java.util.List;
-
-import lombok.RequiredArgsConstructor;
-import jakarta.servlet.http.HttpServletResponse;
-import jakarta.validation.constraints.*;
-import cn.dev33.satoken.annotation.SaCheckPermission;
-import org.ruoyi.domain.bo.knowledge.KnowledgeGraphSegmentBo;
-import org.ruoyi.domain.vo.knowledge.KnowledgeGraphSegmentVo;
-import org.ruoyi.service.knowledge.IKnowledgeGraphSegmentService;
-import org.springframework.web.bind.annotation.*;
-import org.springframework.validation.annotation.Validated;
-import org.ruoyi.common.idempotent.annotation.RepeatSubmit;
-import org.ruoyi.common.log.annotation.Log;
-import org.ruoyi.common.web.core.BaseController;
-import org.ruoyi.common.mybatis.core.page.PageQuery;
-import org.ruoyi.common.core.domain.R;
-import org.ruoyi.common.core.validate.AddGroup;
-import org.ruoyi.common.core.validate.EditGroup;
-import org.ruoyi.common.log.enums.BusinessType;
-import org.ruoyi.common.excel.utils.ExcelUtil;
-import org.ruoyi.common.mybatis.core.page.TableDataInfo;
-
-/**
- * 知识图谱片段
- *
- * @author ageerle
- * @date 2025-12-17
- */
-@Validated
-@RequiredArgsConstructor
-@RestController
-@RequestMapping("/system/graphSegment")
-public class KnowledgeGraphSegmentController extends BaseController {
-
-    private final IKnowledgeGraphSegmentService knowledgeGraphSegmentService;
-
-    /**
-     * 查询知识图谱片段列表
-     */
-    @SaCheckPermission("system:graphSegment:list")
-    @GetMapping("/list")
-    public TableDataInfo<KnowledgeGraphSegmentVo> list(KnowledgeGraphSegmentBo bo, PageQuery pageQuery) {
-        return knowledgeGraphSegmentService.queryPageList(bo, pageQuery);
-    }
-
-    /**
-     * 导出知识图谱片段列表
-     */
-    @SaCheckPermission("system:graphSegment:export")
-    @Log(title = "知识图谱片段", businessType = BusinessType.EXPORT)
-    @PostMapping("/export")
-    public void export(KnowledgeGraphSegmentBo bo, HttpServletResponse response) {
-        List<KnowledgeGraphSegmentVo> list = knowledgeGraphSegmentService.queryList(bo);
-        ExcelUtil.exportExcel(list, "知识图谱片段", KnowledgeGraphSegmentVo.class, response);
-    }
-
-    /**
-     * 获取知识图谱片段详细信息
-     *
-     * @param id 主键
-     */
-    @SaCheckPermission("system:graphSegment:query")
-    @GetMapping("/{id}")
-    public R<KnowledgeGraphSegmentVo> getInfo(@NotNull(message = "主键不能为空")
-                                     @PathVariable Long id) {
-        return R.ok(knowledgeGraphSegmentService.queryById(id));
-    }
-
-    /**
-     * 新增知识图谱片段
-     */
-    @SaCheckPermission("system:graphSegment:add")
-    @Log(title = "知识图谱片段", businessType = BusinessType.INSERT)
-    @RepeatSubmit()
-    @PostMapping()
-    public R<Void> add(@Validated(AddGroup.class) @RequestBody KnowledgeGraphSegmentBo bo) {
-        return toAjax(knowledgeGraphSegmentService.insertByBo(bo));
-    }
-
-    /**
-     * 修改知识图谱片段
-     */
-    @SaCheckPermission("system:graphSegment:edit")
-    @Log(title = "知识图谱片段", businessType = BusinessType.UPDATE)
-    @RepeatSubmit()
-    @PutMapping()
-    public R<Void> edit(@Validated(EditGroup.class) @RequestBody KnowledgeGraphSegmentBo bo) {
-        return toAjax(knowledgeGraphSegmentService.updateByBo(bo));
-    }
-
-    /**
-     * 删除知识图谱片段
-     *
-     * @param ids 主键串
-     */
-    @SaCheckPermission("system:graphSegment:remove")
-    @Log(title = "知识图谱片段", businessType = BusinessType.DELETE)
-    @DeleteMapping("/{ids}")
-    public R<Void> remove(@NotEmpty(message = "主键不能为空")
-                          @PathVariable Long[] ids) {
-        return toAjax(knowledgeGraphSegmentService.deleteWithValidByIds(List.of(ids), true));
-    }
-}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/enums/ChatModeType.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/enums/ChatModeType.java
@@ -15,7 +15,11 @@ public enum ChatModeType {
    DEEP_SEEK("deepseek", "深度求索"),
    QIAN_WEN("qianwen", "通义千问"),
    OPEN_AI("openai", "openai"),
-    PPIO("ppio", "ppio");
+    PPIO("ppio", "ppio"),
+    CUSTOM_API("custom_api", "自定义API"),
+    MINIMAX("minimax", "MiniMax"),
+    XIAOMI("xiaomi", "小米MiMo"),
+    DIFY("dify", "Dify平台");
    private final String code;
    private final String description;

--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/AbstractChatService.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/AbstractChatService.java
@@ -1,9 +1,13 @@
 package org.ruoyi.service.chat;

+import dev.langchain4j.model.chat.ChatModel;
 import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.openai.OpenAiChatModel;
 import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
 import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;

+import java.time.Duration;
+
 /**
 * 聊天消息Service接口
 *
@@ -21,6 +25,23 @@ public interface AbstractChatService {
     */
    StreamingChatModel buildStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest);

+    /**
+     * 创建同步聊天模型（供 Agent/SupervisorAgent 使用）
+     * 默认实现使用 OpenAI 兼容协议，适用于 OpenAI、DeepSeek、PPIO 等兼容接口的 provider。
+     * ZhiPu、QianWen、Ollama 等需覆盖此方法使用各自 SDK。
+     *
+     * @param chatModelVo 模型配置
+     * @return 同步聊天模型实例
+     */
+    default ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        return OpenAiChatModel.builder()
+            .baseUrl(chatModelVo.getApiHost())
+            .apiKey(chatModelVo.getApiKey())
+            .modelName(chatModelVo.getModelName())
+            .timeout(Duration.ofSeconds(120))
+            .build();
+    }
+
    /**
     * 获取服务提供商名称
     */
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/ChatServiceFacade.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/ChatServiceFacade.java
@@ -47,6 +47,7 @@ import org.ruoyi.common.sse.core.SseEmitterManager;
 import org.ruoyi.common.sse.utils.SseMessageUtils;
 import org.ruoyi.domain.bo.vector.QueryVectorBo;
 import org.ruoyi.domain.vo.knowledge.KnowledgeInfoVo;
+import org.ruoyi.enums.ChatModeType;
 import org.ruoyi.factory.ChatServiceFactory;
 import org.ruoyi.mcp.service.core.ToolProviderFactory;
 import org.ruoyi.observability.*;
@@ -97,6 +98,8 @@ public class ChatServiceFacade implements IChatService {

    private final ToolProviderFactory toolProviderFactory;

+    private final org.ruoyi.service.chat.impl.provider.DifyWorkflowService difyWorkflowService;
+
    /**
     * 内存实例缓存，避免同一会话重复创建
     * Key: sessionId, Value: MessageWindowChatMemory实例
@@ -163,6 +166,14 @@ public class ChatServiceFacade implements IChatService {
     * @return 如果需要提前返回则返回SseEmitter，否则返回null
     */
    private SseEmitter handleSpecialChatModes(ChatRequest chatRequest) {
+        // 处理 Dify 工作流对话
+        if (chatRequest.getEnableWorkFlow()
+                && chatRequest.getChatModelVo() != null
+                && ChatModeType.DIFY.getCode().equals(chatRequest.getChatModelVo().getProviderCode())) {
+            log.info("处理Dify工作流对话,会话: {}", chatRequest.getSessionId());
+            return difyWorkflowService.streaming(chatRequest.getChatModelVo(), chatRequest);
+        }
+
        // 处理工作流对话
        if (chatRequest.getEnableWorkFlow()) {
            log.info("处理工作流对话,会话: {}", chatRequest.getSessionId());
@@ -430,8 +441,12 @@ public class ChatServiceFacade implements IChatService {
            }
        }

+        // Dify 自带 RAG 知识库检索，跳过本地向量库查询
+        boolean isDifyProvider = chatRequest.getChatModelVo() != null
+                && ChatModeType.DIFY.getCode().equals(chatRequest.getChatModelVo().getProviderCode());
+
        // 从向量库查询相关历史消息（知识库内容作为上下文）
-        if (chatRequest.getKnowledgeId() != null) {
+        if (chatRequest.getKnowledgeId() != null && !isDifyProvider) {
            // 查询知识库信息
            KnowledgeInfoVo knowledgeInfoVo = knowledgeInfoService.queryById(Long.valueOf(chatRequest.getKnowledgeId()));
            if (knowledgeInfoVo == null) {
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/CustomApiServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/CustomApiServiceImpl.java
@@ -0,0 +1,72 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import cn.hutool.core.util.StrUtil;
+import dev.langchain4j.model.chat.ChatModel;
+import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.openai.OpenAiChatModel;
+import dev.langchain4j.model.openai.OpenAiStreamingChatModel;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
+import org.ruoyi.observability.MyChatModelListener;
+import org.ruoyi.service.chat.AbstractChatService;
+import org.springframework.stereotype.Service;
+
+import java.time.Duration;
+import java.util.List;
+
+/**
+ * 自定义 API 服务调用
+ *
+ * 适用于 OpenAI 兼容接口或仅通过通用 HTTP 协议接入的第三方大模型服务。
+ * 通过模型配置中的 apiHost / apiKey / modelName 即可复用，不需要再写死具体供应商。
+ *
+ * @author better
+ */
+@Service
+@Slf4j
+@RequiredArgsConstructor
+public class CustomApiServiceImpl implements AbstractChatService {
+
+    private static final Duration DEFAULT_TIMEOUT = Duration.ofSeconds(180);
+
+    @Override
+    public StreamingChatModel buildStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest) {
+        return OpenAiStreamingChatModel.builder()
+            .baseUrl(normalizeBaseUrl(chatModelVo.getApiHost()))
+            .apiKey(defaultIfBlank(chatModelVo.getApiKey(), "EMPTY"))
+            .modelName(chatModelVo.getModelName())
+            .timeout(DEFAULT_TIMEOUT)
+            .listeners(List.of(new MyChatModelListener()))
+            .returnThinking(chatRequest.getEnableThinking())
+            .build();
+    }
+
+    @Override
+    public ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        return OpenAiChatModel.builder()
+            .baseUrl(normalizeBaseUrl(chatModelVo.getApiHost()))
+            .apiKey(defaultIfBlank(chatModelVo.getApiKey(), "EMPTY"))
+            .modelName(chatModelVo.getModelName())
+            .timeout(DEFAULT_TIMEOUT)
+            .build();
+    }
+
+    @Override
+    public String getProviderName() {
+        return ChatModeType.CUSTOM_API.getCode();
+    }
+
+    private String normalizeBaseUrl(String baseUrl) {
+        if (StrUtil.isBlank(baseUrl)) {
+            throw new IllegalArgumentException("自定义API的请求地址(apiHost)不能为空");
+        }
+        return baseUrl.endsWith("/") ? baseUrl.substring(0, baseUrl.length() - 1) : baseUrl;
+    }
+
+    private String defaultIfBlank(String value, String defaultValue) {
+        return StrUtil.isBlank(value) ? defaultValue : value;
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyChatServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyChatServiceImpl.java
@@ -0,0 +1,43 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import dev.langchain4j.model.chat.ChatModel;
+import dev.langchain4j.model.chat.StreamingChatModel;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
+import org.ruoyi.service.chat.AbstractChatService;
+import org.ruoyi.service.chat.impl.provider.model.DifyStreamingChatModel;
+import org.springframework.stereotype.Service;
+
+/**
+ * Dify 平台对话服务
+ * <p>
+ * 通过 dify-java-client 接入 Dify 的对话型应用 (Chat App) 和
+ * 工作流编排对话应用 (Chatflow App)，支持流式 SSE 响应。
+ *
+ * @author better
+ */
+@Service
+@Slf4j
+@RequiredArgsConstructor
+public class DifyChatServiceImpl implements AbstractChatService {
+
+    private final DifyConversationService difyConversationService;
+
+    @Override
+    public StreamingChatModel buildStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest) {
+        return new DifyStreamingChatModel(chatModelVo, chatRequest, difyConversationService);
+    }
+
+    @Override
+    public ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        throw new UnsupportedOperationException("Dify 不支持同步 ChatModel，请使用流式模式");
+    }
+
+    @Override
+    public String getProviderName() {
+        return ChatModeType.DIFY.getCode();
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyConversationService.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyConversationService.java
@@ -0,0 +1,35 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import org.springframework.stereotype.Service;
+
+import java.util.concurrent.ConcurrentHashMap;
+
+/**
+ * Dify 会话映射管理
+ * <p>
+ * 维护 ruoyi sessionId 与 Dify conversation_id 的映射关系，
+ * 确保多轮对话上下文连续。
+ *
+ * @author better
+ */
+@Service
+public class DifyConversationService {
+
+    private final ConcurrentHashMap<Long, String> conversationMap = new ConcurrentHashMap<>();
+
+    public String getConversationId(Long sessionId) {
+        return conversationMap.get(sessionId);
+    }
+
+    public void saveMapping(Long sessionId, String difyConversationId) {
+        if (sessionId != null && difyConversationId != null) {
+            conversationMap.put(sessionId, difyConversationId);
+        }
+    }
+
+    public void clearMapping(Long sessionId) {
+        if (sessionId != null) {
+            conversationMap.remove(sessionId);
+        }
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyWorkflowService.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/DifyWorkflowService.java
@@ -0,0 +1,137 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.node.ObjectNode;
+import io.github.imfangs.dify.client.DifyClientFactory;
+import io.github.imfangs.dify.client.DifyWorkflowClient;
+import io.github.imfangs.dify.client.enums.ResponseMode;
+import io.github.imfangs.dify.client.event.ErrorEvent;
+import io.github.imfangs.dify.client.event.WorkflowFinishedEvent;
+import io.github.imfangs.dify.client.event.WorkflowTextChunkEvent;
+import io.github.imfangs.dify.client.callback.WorkflowStreamCallback;
+import io.github.imfangs.dify.client.model.workflow.WorkflowRunRequest;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.dto.request.WorkFlowRunner;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.common.sse.utils.SseMessageUtils;
+import org.springframework.stereotype.Service;
+import org.springframework.web.servlet.mvc.method.annotation.SseEmitter;
+
+import java.util.HashMap;
+import java.util.Iterator;
+import java.util.List;
+import java.util.Map;
+import java.util.concurrent.CompletableFuture;
+
+/**
+ * Dify 工作流执行服务
+ * <p>
+ * 通过 DifyWorkflowClient 调用 Dify 平台上部署的工作流应用，
+ * 并将节点事件通过 SSE 实时推送给前端。
+ *
+ * @author better
+ */
+@Service
+@Slf4j
+public class DifyWorkflowService {
+
+    /**
+     * 流式执行 Dify 工作流
+     *
+     * @param chatModelVo 模型配置（apiHost= Dify 地址, apiKey= Dify 密钥）
+     * @param chatRequest 聊天请求
+     * @return SSE emitter
+     */
+    public SseEmitter streaming(ChatModelVo chatModelVo, ChatRequest chatRequest) {
+        Long userId = chatRequest.getUserId();
+        String tokenValue = chatRequest.getTokenValue();
+        SseEmitter emitter = chatRequest.getEmitter();
+
+        // 构建 Dify 工作流请求参数
+        Map<String, Object> inputs = convertInputs(chatRequest.getWorkFlowRunner());
+
+        WorkflowRunRequest request = WorkflowRunRequest.builder()
+                .inputs(inputs)
+                .responseMode(ResponseMode.STREAMING)
+                .user(String.valueOf(userId))
+                .build();
+
+        DifyWorkflowClient client = DifyClientFactory.createWorkflowClient(
+                normalizeBaseUrl(chatModelVo.getApiHost()),
+                chatModelVo.getApiKey());
+
+        // 异步执行，避免阻塞请求线程
+        CompletableFuture.runAsync(() -> {
+            try {
+                client.runWorkflowStream(request, new WorkflowStreamCallback() {
+
+                    @Override
+                    public void onWorkflowTextChunk(WorkflowTextChunkEvent event) {
+                        String text = event.getData() != null ? event.getData().getText() : null;
+                        if (text != null) {
+                            SseMessageUtils.sendContent(userId, text);
+                        }
+                    }
+
+                    @Override
+                    public void onWorkflowFinished(WorkflowFinishedEvent event) {
+                        // 将最终输出作为内容发送
+                        if (event.getData() != null && event.getData().getOutputs() != null) {
+                            Map<String, Object> outputs = event.getData().getOutputs();
+                            for (Map.Entry<String, Object> entry : outputs.entrySet()) {
+                                SseMessageUtils.sendContent(userId,
+                                        entry.getKey() + ": " + entry.getValue() + "\n");
+                            }
+                        }
+                        SseMessageUtils.sendDone(userId);
+                        SseMessageUtils.completeConnection(userId, tokenValue);
+                    }
+
+                    @Override
+                    public void onError(ErrorEvent event) {
+                        SseMessageUtils.sendError(userId, event.getMessage());
+                    }
+
+                    @Override
+                    public void onException(Throwable throwable) {
+                        log.error("Dify 工作流执行异常", throwable);
+                        SseMessageUtils.sendError(userId, throwable.getMessage());
+                        SseMessageUtils.completeConnection(userId, tokenValue);
+                    }
+                });
+            } catch (Exception e) {
+                log.error("Dify 工作流执行失败", e);
+                SseMessageUtils.sendError(userId, e.getMessage());
+                SseMessageUtils.completeConnection(userId, tokenValue);
+            }
+        });
+
+        return emitter;
+    }
+
+    /**
+     * 将 WorkFlowRunner.inputs (List<ObjectNode>) 转换为 Dify 所需的 Map
+     */
+    private Map<String, Object> convertInputs(WorkFlowRunner runner) {
+        Map<String, Object> result = new HashMap<>();
+        if (runner == null || runner.getInputs() == null) {
+            return result;
+        }
+        for (ObjectNode node : runner.getInputs()) {
+            Iterator<Map.Entry<String, JsonNode>> fields = node.fields();
+            while (fields.hasNext()) {
+                Map.Entry<String, JsonNode> field = fields.next();
+                result.put(field.getKey(), field.getValue().asText());
+            }
+        }
+        return result;
+    }
+
+    private String normalizeBaseUrl(String baseUrl) {
+        if (baseUrl == null || baseUrl.isBlank()) {
+            throw new IllegalArgumentException("Dify API 地址(apiHost)不能为空");
+        }
+        return baseUrl.endsWith("/") ? baseUrl.substring(0, baseUrl.length() - 1) : baseUrl;
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/MiMoServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/MiMoServiceImpl.java
@@ -0,0 +1,47 @@
+package org.ruoyi.service.chat.impl.provider;
+
+
+import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.openai.OpenAiStreamingChatModel;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
+import org.ruoyi.observability.MyChatModelListener;
+import org.ruoyi.service.chat.AbstractChatService;
+import org.springframework.stereotype.Service;
+
+import java.util.List;
+
+
+/**
+ * 小米MiMo服务调用
+ * <p>
+ * 小米提供OpenAI兼容的API接口，支持MiMo等模型。
+ *
+ * @author ageerle
+ * @date 2026/4/19
+ */
+@Service
+@Slf4j
+@RequiredArgsConstructor
+public class MiMoServiceImpl implements AbstractChatService {
+
+    @Override
+    public StreamingChatModel buildStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest) {
+        return OpenAiStreamingChatModel.builder()
+            .baseUrl(chatModelVo.getApiHost())
+            .apiKey(chatModelVo.getApiKey())
+            .modelName(chatModelVo.getModelName())
+            .listeners(List.of(new MyChatModelListener()))
+            .returnThinking(chatRequest.getEnableThinking())
+            .build();
+    }
+
+    @Override
+    public String getProviderName() {
+        return ChatModeType.XIAOMI.getCode();
+    }
+
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/MinimaxServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/MinimaxServiceImpl.java
@@ -0,0 +1,44 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.openai.OpenAiStreamingChatModel;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
+import org.ruoyi.observability.MyChatModelListener;
+import org.ruoyi.service.chat.AbstractChatService;
+import org.springframework.stereotype.Service;
+
+import java.util.List;
+
+/**
+ * MiniMax服务调用
+ * <p>
+ * MiniMax提供OpenAI兼容的API接口，支持MiniMax-M2.7、MiniMax-M2.5等模型。
+ * API地址：https://api.minimax.io/v1
+ *
+ * @author octopus
+ * @date 2026/3/21
+ */
+@Service
+@Slf4j
+public class MinimaxServiceImpl implements AbstractChatService {
+
+    @Override
+    public StreamingChatModel buildStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest) {
+        return OpenAiStreamingChatModel.builder()
+            .baseUrl(chatModelVo.getApiHost())
+            .apiKey(chatModelVo.getApiKey())
+            .modelName(chatModelVo.getModelName())
+            .listeners(List.of(new MyChatModelListener()))
+            .returnThinking(chatRequest.getEnableThinking())
+            .build();
+    }
+
+    @Override
+    public String getProviderName() {
+        return ChatModeType.MINIMAX.getCode();
+    }
+
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/OllamaServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/OllamaServiceImpl.java
@@ -1,7 +1,9 @@
 package org.ruoyi.service.chat.impl.provider;


+import dev.langchain4j.model.chat.ChatModel;
 import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.ollama.OllamaChatModel;
 import dev.langchain4j.model.ollama.OllamaStreamingChatModel;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -37,6 +39,14 @@ public class OllamaServiceImpl implements AbstractChatService {
                .build();
    }

+    @Override
+    public ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        return OllamaChatModel.builder()
+            .baseUrl(chatModelVo.getApiHost())
+            .modelName(chatModelVo.getModelName())
+            .build();
+    }
+
    @Override
    public String getProviderName() {
        return ChatModeType.OLLAMA.getCode();
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/QianWenChatServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/QianWenChatServiceImpl.java
@@ -1,7 +1,9 @@
 package org.ruoyi.service.chat.impl.provider;


+import dev.langchain4j.community.model.dashscope.QwenChatModel;
 import dev.langchain4j.community.model.dashscope.QwenStreamingChatModel;
+import dev.langchain4j.model.chat.ChatModel;
 import dev.langchain4j.model.chat.StreamingChatModel;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -38,6 +40,14 @@ public class QianWenChatServiceImpl implements AbstractChatService {
                .build();
    }

+    @Override
+    public ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        return QwenChatModel.builder()
+            .apiKey(chatModelVo.getApiKey())
+            .modelName(chatModelVo.getModelName())
+            .build();
+    }
+
    @Override
    public String getProviderName() {
        return ChatModeType.QIAN_WEN.getCode();
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/ZhiPuChatServiceImpl.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/ZhiPuChatServiceImpl.java
@@ -1,7 +1,9 @@
 package org.ruoyi.service.chat.impl.provider;


+import dev.langchain4j.community.model.zhipu.ZhipuAiChatModel;
 import dev.langchain4j.community.model.zhipu.ZhipuAiStreamingChatModel;
+import dev.langchain4j.model.chat.ChatModel;
 import dev.langchain4j.model.chat.StreamingChatModel;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -35,6 +37,14 @@ public class ZhiPuChatServiceImpl implements AbstractChatService {
            .build();
    }

+    @Override
+    public ChatModel buildChatModel(ChatModelVo chatModelVo) {
+        return ZhipuAiChatModel.builder()
+            .apiKey(chatModelVo.getApiKey())
+            .model(chatModelVo.getModelName())
+            .build();
+    }
+
    @Override
    public String getProviderName() {
        return ChatModeType.ZHI_PU.getCode();
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/model/DifyStreamingChatModel.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/chat/impl/provider/model/DifyStreamingChatModel.java
@@ -0,0 +1,172 @@
+package org.ruoyi.service.chat.impl.provider.model;
+
+import dev.langchain4j.data.message.AiMessage;
+import dev.langchain4j.data.message.ChatMessage;
+import dev.langchain4j.data.message.UserMessage;
+import dev.langchain4j.model.chat.StreamingChatModel;
+import dev.langchain4j.model.chat.response.ChatResponse;
+import dev.langchain4j.model.chat.response.StreamingChatResponseHandler;
+import io.github.imfangs.dify.client.DifyChatClient;
+import io.github.imfangs.dify.client.DifyClientFactory;
+import io.github.imfangs.dify.client.enums.ResponseMode;
+import io.github.imfangs.dify.client.event.ErrorEvent;
+import io.github.imfangs.dify.client.event.MessageEndEvent;
+import io.github.imfangs.dify.client.event.MessageEvent;
+import lombok.extern.slf4j.Slf4j;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.service.chat.impl.provider.DifyConversationService;
+
+import java.util.List;
+
+/**
+ * Dify 流式聊天模型适配器
+ * <p>
+ * 将 Dify 的回调式流式响应适配为 langchain4j 的 StreamingChatModel 接口，
+ * 使 ChatServiceFacade 可以像其他 provider 一样统一调用。
+ *
+ * @author better
+ */
+@Slf4j
+public class DifyStreamingChatModel implements StreamingChatModel {
+
+    private final ChatModelVo chatModelVo;
+    private final ChatRequest chatRequest;
+    private final DifyConversationService conversationService;
+
+    public DifyStreamingChatModel(ChatModelVo chatModelVo, ChatRequest chatRequest,
+                                  DifyConversationService conversationService) {
+        this.chatModelVo = chatModelVo;
+        this.chatRequest = chatRequest;
+        this.conversationService = conversationService;
+    }
+
+    @Override
+    public void chat(List<ChatMessage> messages, StreamingChatResponseHandler handler) {
+        // 1. 从 langchain4j 消息列表中提取最后一条用户消息作为 query
+        String query = extractUserQuery(messages);
+
+        // 2. 获取 Dify conversation_id（多轮对话连续性）
+        String conversationId = null;
+        if (chatRequest.getSessionId() != null) {
+            conversationId = conversationService.getConversationId(chatRequest.getSessionId());
+        }
+
+        // 3. 构建 Dify 请求
+        io.github.imfangs.dify.client.model.chat.ChatMessage difyMessage = io.github.imfangs.dify.client.model.chat.ChatMessage.builder()
+                .query(query)
+                .user(String.valueOf(chatRequest.getUserId()))
+                .responseMode(ResponseMode.STREAMING)
+                .conversationId(conversationId)
+                .autoGenerateName(true)
+                .build();
+
+        // 4. 创建 Dify 客户端并发送流式请求
+        try {
+            DifyChatClient client = DifyClientFactory.createChatClient(
+                    normalizeBaseUrl(chatModelVo.getApiHost()),
+                    chatModelVo.getApiKey());
+
+            client.sendChatMessageStream(difyMessage, new DifyChatStreamAdapter(handler));
+        } catch (Exception e) {
+            log.error("Dify 流式对话调用失败", e);
+            handler.onError(e);
+        }
+    }
+
+    @Override
+    public void chat(String userMessage, StreamingChatResponseHandler handler) {
+        io.github.imfangs.dify.client.model.chat.ChatMessage difyMessage = io.github.imfangs.dify.client.model.chat.ChatMessage.builder()
+                .query(userMessage)
+                .user(String.valueOf(chatRequest.getUserId()))
+                .responseMode(ResponseMode.STREAMING)
+                .conversationId(chatRequest.getSessionId() != null
+                        ? conversationService.getConversationId(chatRequest.getSessionId()) : null)
+                .autoGenerateName(true)
+                .build();
+
+        try {
+            DifyChatClient client = DifyClientFactory.createChatClient(
+                    normalizeBaseUrl(chatModelVo.getApiHost()),
+                    chatModelVo.getApiKey());
+
+            client.sendChatMessageStream(difyMessage, new DifyChatStreamAdapter(handler));
+        } catch (Exception e) {
+            log.error("Dify 流式对话调用失败", e);
+            handler.onError(e);
+        }
+    }
+
+    /**
+     * 从 langchain4j 消息列表中提取最后一条用户消息文本
+     */
+    private String extractUserQuery(List<ChatMessage> messages) {
+        for (int i = messages.size() - 1; i >= 0; i--) {
+            ChatMessage msg = messages.get(i);
+            if (msg instanceof UserMessage) {
+                return ((UserMessage) msg).singleText();
+            }
+        }
+        return "";
+    }
+
+    private String normalizeBaseUrl(String baseUrl) {
+        if (baseUrl == null || baseUrl.isBlank()) {
+            throw new IllegalArgumentException("Dify API 地址(apiHost)不能为空");
+        }
+        return baseUrl.endsWith("/") ? baseUrl.substring(0, baseUrl.length() - 1) : baseUrl;
+    }
+
+    /**
+     * Dify 回调适配器
+     * 将 Dify ChatStreamCallback 事件转发给 langchain4j StreamingChatResponseHandler
+     */
+    private class DifyChatStreamAdapter implements io.github.imfangs.dify.client.callback.ChatStreamCallback {
+
+        private final StreamingChatResponseHandler handler;
+        private final StringBuilder fullResponse = new StringBuilder();
+
+        DifyChatStreamAdapter(StreamingChatResponseHandler handler) {
+            this.handler = handler;
+        }
+
+        @Override
+        public void onMessage(MessageEvent event) {
+            String answer = event.getAnswer();
+            if (answer != null) {
+                fullResponse.append(answer);
+                handler.onPartialResponse(answer);
+            }
+            // 保存 Dify conversation_id 以维持多轮对话
+            if (event.getConversationId() != null && chatRequest.getSessionId() != null) {
+                conversationService.saveMapping(chatRequest.getSessionId(), event.getConversationId());
+            }
+        }
+
+        @Override
+        public void onMessageEnd(MessageEndEvent event) {
+            // 保存 conversation_id
+            if (event.getConversationId() != null && chatRequest.getSessionId() != null) {
+                conversationService.saveMapping(chatRequest.getSessionId(), event.getConversationId());
+            }
+
+            // 构建完整的 ChatResponse 交给上层处理
+            AiMessage aiMessage = new AiMessage(fullResponse.toString());
+            ChatResponse response = ChatResponse.builder()
+                    .aiMessage(aiMessage)
+                    .id(event.getMessageId())
+                    .build();
+            handler.onCompleteResponse(response);
+        }
+
+        @Override
+        public void onError(ErrorEvent event) {
+            handler.onError(new RuntimeException(event.getMessage()));
+        }
+
+        @Override
+        public void onException(Throwable throwable) {
+            handler.onError(throwable);
+        }
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/embed/impl/MinimaxEmbeddingProvider.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/embed/impl/MinimaxEmbeddingProvider.java
@@ -0,0 +1,17 @@
+package org.ruoyi.service.embed.impl;
+
+import org.springframework.stereotype.Component;
+
+/**
+ * MiniMax嵌入模型（兼容OpenAI接口）
+ * <p>
+ * 支持embo-01模型，1536维度向量。
+ * API地址：https://api.minimax.io/v1
+ *
+ * @author octopus
+ * @date 2026/3/21
+ */
+@Component("minimax")
+public class MinimaxEmbeddingProvider extends OpenAiEmbeddingProvider {
+
+}
--- a/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/embed/impl/ZhipuAiEmbeddingProvider.java
+++ b/ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/service/embed/impl/ZhipuAiEmbeddingProvider.java
@@ -0,0 +1,48 @@
+package org.ruoyi.service.embed.impl;
+
+
+
+import dev.langchain4j.community.model.zhipu.ZhipuAiEmbeddingModel;
+import dev.langchain4j.data.embedding.Embedding;
+import dev.langchain4j.data.segment.TextSegment;
+import dev.langchain4j.model.embedding.EmbeddingModel;
+import dev.langchain4j.model.output.Response;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ModalityType;
+import org.ruoyi.service.embed.BaseEmbedModelService;
+import org.springframework.stereotype.Component;
+
+import java.util.List;
+import java.util.Set;
+
+/**
+ * @Author:yang
+ * @Date:
+ * @Description: 智谱AI嵌入模型
+ */
+@Component("zhipu")
+public class ZhipuAiEmbeddingProvider implements BaseEmbedModelService {
+    protected ChatModelVo chatModelVo;
+
+    @Override
+    public void configure(ChatModelVo config) {
+        this.chatModelVo = config;
+    }
+
+    @Override
+    public Set<ModalityType> getSupportedModalities() {
+        return Set.of(ModalityType.TEXT);
+    }
+
+    @Override
+    public Response<List<Embedding>> embedAll(List<TextSegment> textSegments) {
+        EmbeddingModel model = ZhipuAiEmbeddingModel.builder()
+            .baseUrl(chatModelVo.getApiHost())
+            .apiKey(chatModelVo.getApiKey())
+            .model(chatModelVo.getModelName())
+            .dimensions(chatModelVo.getModelDimension())
+            .build();
+
+        return model.embedAll(textSegments);
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/enums/ChatModeTypeTest.java
+++ b/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/enums/ChatModeTypeTest.java
@@ -0,0 +1,42 @@
+package org.ruoyi.enums;
+
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+/**
+ * Unit tests for ChatModeType enum
+ */
+class ChatModeTypeTest {
+
+    @Test
+    void minimaxEnumExists() {
+        ChatModeType minimax = ChatModeType.MINIMAX;
+        assertNotNull(minimax);
+    }
+
+    @Test
+    void minimaxCode_isMinimax() {
+        assertEquals("minimax", ChatModeType.MINIMAX.getCode());
+    }
+
+    @Test
+    void minimaxDescription_isMiniMax() {
+        assertEquals("MiniMax", ChatModeType.MINIMAX.getDescription());
+    }
+
+    @Test
+    void allProviders_haveUniqueCode() {
+        ChatModeType[] values = ChatModeType.values();
+        long uniqueCodes = java.util.Arrays.stream(values)
+            .map(ChatModeType::getCode)
+            .distinct()
+            .count();
+        assertEquals(values.length, uniqueCodes, "All providers must have unique codes");
+    }
+
+    @Test
+    void valueOf_minimax() {
+        assertEquals(ChatModeType.MINIMAX, ChatModeType.valueOf("MINIMAX"));
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/integration/MinimaxIntegrationTest.java
+++ b/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/integration/MinimaxIntegrationTest.java
@@ -0,0 +1,70 @@
+package org.ruoyi.integration;
+
+import dev.langchain4j.model.chat.StreamingChatModel;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.api.condition.EnabledIfEnvironmentVariable;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.service.chat.impl.provider.MinimaxServiceImpl;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+/**
+ * Integration tests for MiniMax provider.
+ * These tests require a valid MINIMAX_API_KEY environment variable.
+ */
+@EnabledIfEnvironmentVariable(named = "MINIMAX_API_KEY", matches = ".+")
+class MinimaxIntegrationTest {
+
+    private MinimaxServiceImpl minimaxService;
+    private String apiKey;
+
+    @BeforeEach
+    void setUp() {
+        minimaxService = new MinimaxServiceImpl();
+        apiKey = System.getenv("MINIMAX_API_KEY");
+    }
+
+    @Test
+    void buildStreamingChatModel_withRealApiKey_M27() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey(apiKey);
+        modelVo.setModelName("MiniMax-M2.7");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(false);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model, "Should create streaming model with real API key");
+    }
+
+    @Test
+    void buildStreamingChatModel_withRealApiKey_M25() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey(apiKey);
+        modelVo.setModelName("MiniMax-M2.5");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(false);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model, "Should create streaming model with M2.5");
+    }
+
+    @Test
+    void buildStreamingChatModel_withRealApiKey_M25Highspeed() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey(apiKey);
+        modelVo.setModelName("MiniMax-M2.5-highspeed");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(false);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model, "Should create streaming model with M2.5-highspeed");
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/service/chat/impl/provider/MinimaxServiceImplTest.java
+++ b/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/service/chat/impl/provider/MinimaxServiceImplTest.java
@@ -0,0 +1,76 @@
+package org.ruoyi.service.chat.impl.provider;
+
+import dev.langchain4j.model.chat.StreamingChatModel;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.ruoyi.common.chat.domain.dto.request.ChatRequest;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ChatModeType;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+/**
+ * Unit tests for MinimaxServiceImpl
+ */
+class MinimaxServiceImplTest {
+
+    private MinimaxServiceImpl minimaxService;
+
+    @BeforeEach
+    void setUp() {
+        minimaxService = new MinimaxServiceImpl();
+    }
+
+    @Test
+    void getProviderName_returnsMinimaxCode() {
+        assertEquals("minimax", minimaxService.getProviderName());
+        assertEquals(ChatModeType.MINIMAX.getCode(), minimaxService.getProviderName());
+    }
+
+    @Test
+    void buildStreamingChatModel_returnsNonNull() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey("test-api-key");
+        modelVo.setModelName("MiniMax-M2.7");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(false);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model);
+    }
+
+    @Test
+    void buildStreamingChatModel_withThinkingEnabled() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey("test-api-key");
+        modelVo.setModelName("MiniMax-M2.5");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(true);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model);
+    }
+
+    @Test
+    void buildStreamingChatModel_withHighspeedModel() {
+        ChatModelVo modelVo = new ChatModelVo();
+        modelVo.setApiHost("https://api.minimax.io/v1");
+        modelVo.setApiKey("test-api-key");
+        modelVo.setModelName("MiniMax-M2.5-highspeed");
+
+        ChatRequest request = new ChatRequest();
+        request.setEnableThinking(false);
+
+        StreamingChatModel model = minimaxService.buildStreamingChatModel(modelVo, request);
+        assertNotNull(model);
+    }
+
+    @Test
+    void implementsAbstractChatService() {
+        assertInstanceOf(org.ruoyi.service.chat.AbstractChatService.class, minimaxService);
+    }
+}
--- a/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/service/embed/impl/MinimaxEmbeddingProviderTest.java
+++ b/ruoyi-modules/ruoyi-chat/src/test/java/org/ruoyi/service/embed/impl/MinimaxEmbeddingProviderTest.java
@@ -0,0 +1,55 @@
+package org.ruoyi.service.embed.impl;
+
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.ruoyi.common.chat.domain.vo.chat.ChatModelVo;
+import org.ruoyi.enums.ModalityType;
+import org.ruoyi.service.embed.BaseEmbedModelService;
+
+import java.util.Set;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+/**
+ * Unit tests for MinimaxEmbeddingProvider
+ */
+class MinimaxEmbeddingProviderTest {
+
+    private MinimaxEmbeddingProvider provider;
+
+    @BeforeEach
+    void setUp() {
+        provider = new MinimaxEmbeddingProvider();
+    }
+
+    @Test
+    void implementsBaseEmbedModelService() {
+        assertInstanceOf(BaseEmbedModelService.class, provider);
+    }
+
+    @Test
+    void extendsOpenAiEmbeddingProvider() {
+        assertInstanceOf(OpenAiEmbeddingProvider.class, provider);
+    }
+
+    @Test
+    void getSupportedModalities_returnsText() {
+        Set<ModalityType> modalities = provider.getSupportedModalities();
+        assertNotNull(modalities);
+        assertTrue(modalities.contains(ModalityType.TEXT));
+        assertEquals(1, modalities.size());
+    }
+
+    @Test
+    void configure_setsModelConfig() {
+        ChatModelVo config = new ChatModelVo();
+        config.setApiHost("https://api.minimax.io/v1");
+        config.setApiKey("test-api-key");
+        config.setModelName("embo-01");
+        config.setModelDimension(1536);
+
+        provider.configure(config);
+        // configure sets internal state; verify no exception thrown
+        assertNotNull(provider);
+    }
+}
Author	SHA1	Message	Date
wangle	bc151e49c5	feat: 添加Dify平台集成支持 - 升级 dify-sdk-java 从 1.0.7 到 1.2.6 - 新增 ChatModeType.DIFY 枚举类型 - 新增 DifyChatServiceImpl、DifyConversationService、DifyWorkflowService 实现 - 新增 DifyStreamingChatModel 流式聊天模型 - 支持Dify工作流对话模式 - Dify自带RAG知识库时跳过本地向量库查询 Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-04-23 19:19:26 +08:00
wangle	2c6ff66830	fix: 修正application.yml演示模式message缩进 Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-04-19 13:42:39 +08:00
wangle	4f79a66559	feat: 添加小米MiMo、DeepSeek、自定义厂商等provider支持 - 新增小米MiMo服务实现类(MiMoServiceImpl) - ChatModeType添加XIAOMI枚举 - 更新SQL初始化脚本，新增多家厂商(provider)和模型数据 - 添加2026-04-19数据库更新脚本 - application.yml演示模式排除路径增加attach/fragment/info接口 - 删除独立的minimax_provider.sql(数据已合并到主SQL) Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-04-19 13:42:05 +08:00
wangle	22883b4334	Merge branch pr-280: 添加MiniMax作为LLM提供商解决冲突: - README: 保留Qdrant向量库信息 + 合并MiniMax模型接入 - pom.xml: 保留spring-boot-starter-test + 添加MiniMax测试依赖 - ChatModeType: 保留CUSTOM_API + 新增MINIMAX枚举 - MinimaxServiceImpl: 保留MyChatModelListener监听器 Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-04-19 11:47:04 +08:00
wangle	081da6d18d	feat: 添加MiniMax作为LLM提供商，合并PR#280并补充监听合并PR#280的MiniMax provider实现，解决与main分支的冲突，并在MinimaxServiceImpl中补充MyChatModelListener监听，与其他provider保持一致。 Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-04-17 18:31:53 +08:00
ageerle	74eb5b2530	Merge pull request #290 from yangzhen233/feature/zhipu-embedding 添加智谱向量模型实现	2026-04-17 18:24:24 +08:00
ageerle	b0328fe0ef	Merge pull request #291 from xiaonieli7/main fate：增加自定义模型，调整前端模型选择下拉框	2026-04-17 18:16:59 +08:00
Administrator	2ee0aae57e	fate：增加自定义模型，调整前端模型选择下拉框	2026-04-17 08:31:40 +08:00
杨振	d9c3de660a	添加智谱向量模型实现	2026-04-16 21:18:11 +08:00
ageerle	c4f7c1f5d0	Merge pull request #283 from MrWws/fix/docker-compose-all.yaml fix: 修复运行docker-compose-all.yaml报错的问题	2026-04-14 14:36:14 +08:00
MrWws	b9097b4989	fix: 修复运行docker-compose-all.yaml报错的问题	2026-03-31 22:59:26 +08:00
octopus	5d14eb20af	feat: add MiniMax as first-class LLM provider Add MiniMax AI as the 7th LLM provider, supporting chat (M2.7, M2.5, M2.5-highspeed) and embedding (embo-01) models via OpenAI-compatible API. Changes: - Add MINIMAX enum to ChatModeType - Add MinimaxServiceImpl chat provider (OpenAI-compat streaming) - Add MinimaxEmbeddingProvider for vector embeddings - Add SQL migration for provider and model registration - Add 14 unit tests + 3 integration tests - Update README/README_EN with MiniMax in provider list	2026-03-21 16:14:19 +08:00