diff --git a/config.json b/config.json
index 5a6fdef..8f6d29a 100644
--- a/config.json
+++ b/config.json
@@ -69,6 +69,16 @@
         },
         "server_url":"https://api.minimax.chat/v1/text/chatcompletion_pro"
       }
+    ],
+    "jiutian": [
+      {
+        "models": ["jiutian-qianwen"],
+        "enabled": true,
+        "credentials": {
+          "api_key": "xxx"
+        },
+        "server_url":"https://jiutian.10086.cn/largemodel/api/v1"
+      }
     ]
   }
 }
diff --git a/go.mod b/go.mod
index 069b0d9..10e66b8 100644
--- a/go.mod
+++ b/go.mod
@@ -49,6 +49,7 @@ require (
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-playground/validator/v10 v10.20.0 // indirect
 	github.com/goccy/go-json v0.10.3 // indirect
+	github.com/golang-jwt/jwt/v5 v5.2.1 // indirect
 	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
 	github.com/golang/protobuf v1.5.4 // indirect
 	github.com/google/s2a-go v0.1.7 // indirect
diff --git a/go.sum b/go.sum
index 52452fd..5420cea 100644
--- a/go.sum
+++ b/go.sum
@@ -70,6 +70,8 @@ github.com/go-playground/validator/v10 v10.20.0 h1:K9ISHbSaI0lyB2eWMPJo+kOS/FBEx
 github.com/go-playground/validator/v10 v10.20.0/go.mod h1:dbuPbCMFw/DrkbEynArYaCwl3amGuJotoKCe95atGMM=
 github.com/goccy/go-json v0.10.3 h1:KZ5WoDbxAIgm2HNbYckL0se1fHD6rz5j4ywS6ebzDqA=
 github.com/goccy/go-json v0.10.3/go.mod h1:oq7eo15ShAhp70Anwd5lgX2pLfOS3QCiwU/PULtXL6M=
+github.com/golang-jwt/jwt/v5 v5.2.1 h1:OuVbFODueb089Lh128TAcimifWaLhJwVflnrgM17wHk=
+github.com/golang-jwt/jwt/v5 v5.2.1/go.mod h1:pqrtFR0X4osieyHYxtmOUWsAWrfe1Q5UVIyoH402zdk=
 github.com/golang/glog v0.0.0-20160126235308-23def4e6c14b/go.mod h1:SBH7ygxi8pfUlaOkMMuAQtPIUF8ecWP5IEl/CR7VP2Q=
 github.com/golang/groupcache v0.0.0-20200121045136-8c9f03a8e57e/go.mod h1:cIg4eruTrX1D+g88fzRXU5OdNfaM+9IcxsU14FzY7Hc=
 github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da h1:oI5xCqsCo564l8iNU+DwB5epxmsaqB+rhGL0m5jtYqE=
diff --git a/pkg/adapter/jiutian_openai.go b/pkg/adapter/jiutian_openai.go
new file mode 100644
index 0000000..e17505b
--- /dev/null
+++ b/pkg/adapter/jiutian_openai.go
@@ -0,0 +1,169 @@
+package adapter
+
+import (
+	"github.com/sashabaranov/go-openai"
+	"go.uber.org/zap"
+	"simple-one-api/pkg/llm/jiutian"
+	"simple-one-api/pkg/mylog"
+	"time"
+)
+
+// convertFinishReason 将字符串转换为OpenAI的FinishReason类型
+func convertFinishReason(reason string) openai.FinishReason {
+	switch reason {
+	case "stop":
+		return openai.FinishReasonStop
+	case "length":
+		return openai.FinishReasonLength
+	case "content_filter":
+		return openai.FinishReasonContentFilter
+	case "function_call":
+		return openai.FinishReasonFunctionCall
+	default:
+		return openai.FinishReasonNull
+	}
+}
+
+// OpenAIRequestToJiuTianRequest 将OpenAI请求转换为九天模型请求
+func OpenAIRequestToJiuTianRequest(oaiReq *openai.ChatCompletionRequest) *jiutian.ChatCompletionRequest {
+	mylog.Logger.Info("Converting OpenAI request to JiuTian request",
+		zap.String("model", oaiReq.Model),
+		zap.Int("message_count", len(oaiReq.Messages)),
+		zap.Float32("temperature", oaiReq.Temperature),
+		zap.Float32("top_p", oaiReq.TopP))
+
+	// 获取最后一条消息作为prompt
+	lastMessage := oaiReq.Messages[len(oaiReq.Messages)-1].Content
+	
+	// 构建历史消息
+	var history [][]string
+	if len(oaiReq.Messages) > 1 {
+		for i := 0; i < len(oaiReq.Messages)-1; i += 2 {
+			if i+1 < len(oaiReq.Messages) {
+				history = append(history, []string{
+					oaiReq.Messages[i].Content,
+					oaiReq.Messages[i+1].Content,
+				})
+			}
+		}
+	}
+
+	mylog.Logger.Debug("Request conversion details",
+		zap.String("prompt", lastMessage),
+		zap.Int("history_length", len(history)))
+
+	// 创建请求
+	req := jiutian.NewChatCompletionRequest().
+		WithModelID(oaiReq.Model). // 使用传入的模型ID
+		WithPrompt(lastMessage).
+		WithHistory(history).
+		WithStream(oaiReq.Stream)
+
+	// 设置温度参数（如果有）
+	if oaiReq.Temperature > 0 {
+		req.WithTemperature(oaiReq.Temperature)
+	}
+
+	// 设置top_p参数（如果有）
+	if oaiReq.TopP > 0 {
+		req.WithTopP(oaiReq.TopP)
+	}
+
+	mylog.Logger.Debug("Created JiuTian request",
+		zap.String("model_id", req.ModelID),
+		zap.Bool("stream", req.Stream),
+		zap.Float32("temperature", req.Params.Temperature),
+		zap.Float32("top_p", req.Params.TopP))
+
+	return req
+}
+
+// JiuTianResponseToOpenAIResponse 将九天模型响应转换为OpenAI响应
+func JiuTianResponseToOpenAIResponse(jiutianResp *jiutian.ChatCompletionResponse) *openai.ChatCompletionResponse {
+	mylog.Logger.Info("Converting JiuTian response to OpenAI response",
+		zap.Any("jiutian_response", map[string]interface{}{
+			"usage":    jiutianResp.Usage,
+			"response": jiutianResp.Response,
+			"delta":    jiutianResp.Delta,
+			"finished": jiutianResp.Finished,
+			"history":  jiutianResp.History,
+		}))
+
+	// 创建选项
+	choice := openai.ChatCompletionChoice{
+		Index: 0,
+		Message: openai.ChatCompletionMessage{
+			Role:    "assistant",
+			Content: jiutianResp.Response,
+		},
+	}
+
+	// 设置结束原因
+	if jiutianResp.Finished != "" {
+		choice.FinishReason = convertFinishReason(jiutianResp.Finished)
+	}
+
+	resp := &openai.ChatCompletionResponse{
+		ID:      "jiutian-" + time.Now().Format("20060102150405"),
+		Object:  "chat.completion",
+		Created: time.Now().Unix(),
+		Model:   "Llama3.1-70B", // 使用实际的模型ID
+		Choices: []openai.ChatCompletionChoice{choice},
+		Usage: openai.Usage{
+			PromptTokens:     jiutianResp.Usage.PromptTokens,
+			CompletionTokens: jiutianResp.Usage.CompletionTokens,
+			TotalTokens:      jiutianResp.Usage.TotalTokens,
+		},
+	}
+
+	mylog.Logger.Info("Converted to OpenAI response",
+		zap.Any("openai_response", map[string]interface{}{
+			"id":      resp.ID,
+			"model":   resp.Model,
+			"choices": resp.Choices,
+			"usage":   resp.Usage,
+		}))
+
+	return resp
+}
+
+// JiuTianStreamResponseToOpenAIStreamResponse 将九天模型的流式响应转换为OpenAI流式响应
+func JiuTianStreamResponseToOpenAIStreamResponse(jiutianResp *jiutian.ChatCompletionStreamResponse) *openai.ChatCompletionStreamResponse {
+	mylog.Logger.Info("Converting JiuTian stream response to OpenAI stream response",
+		zap.Any("jiutian_stream_response", map[string]interface{}{
+			"response": jiutianResp.Response,
+			"delta":    jiutianResp.Delta,
+			"finished": jiutianResp.Finished,
+			"history":  jiutianResp.History,
+		}))
+
+	choice := openai.ChatCompletionStreamChoice{
+		Index: 0,
+		Delta: openai.ChatCompletionStreamChoiceDelta{
+			Role:    "assistant",
+			Content: jiutianResp.Response,
+		},
+	}
+
+	// 只在收到结束标记时设置结束原因
+	if jiutianResp.Delta == "[EOS]" {
+		choice.FinishReason = convertFinishReason(jiutianResp.Finished)
+	}
+
+	resp := &openai.ChatCompletionStreamResponse{
+		ID:      "jiutian-stream-" + time.Now().Format("20060102150405"),
+		Object:  "chat.completion.chunk",
+		Created: time.Now().Unix(),
+		Model:   "Llama3.1-70B", // 使用实际的模型ID
+		Choices: []openai.ChatCompletionStreamChoice{choice},
+	}
+
+	mylog.Logger.Info("Converted to OpenAI stream response",
+		zap.Any("openai_stream_response", map[string]interface{}{
+			"id":      resp.ID,
+			"model":   resp.Model,
+			"choices": resp.Choices,
+		}))
+
+	return resp
+} 
\ No newline at end of file
diff --git a/pkg/handler/openai_handler.go b/pkg/handler/openai_handler.go
index b62f5a8..8fb63e2 100644
--- a/pkg/handler/openai_handler.go
+++ b/pkg/handler/openai_handler.go
@@ -4,9 +4,6 @@ import (
 	"bytes"
 	"context"
 	"errors"
-	"github.com/gin-gonic/gin"
-	"github.com/sashabaranov/go-openai"
-	"go.uber.org/zap"
 	"io"
 	"net/http"
 	"simple-one-api/pkg/adapter"
@@ -17,6 +14,10 @@ import (
 	"simple-one-api/pkg/utils"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
+	"github.com/sashabaranov/go-openai"
+	"go.uber.org/zap"
 )
 
 var defaultReqTimeout = 10
@@ -52,6 +53,7 @@ var serviceHandlerMap = map[string]func(*gin.Context, *OAIRequestParam) error{
 	"claude":       OpenAI2ClaudeHandler,
 	"agentbuilder": OpenAI2AgentBuilderHandler,
 	"dify":         OpenAI2DifyHandler,
+	"jiutian":      OpenAI2JiuTianHandler,
 }
 
 func LogRequestDetails(c *gin.Context) {
diff --git a/pkg/handler/openai_jiutian_handler.go b/pkg/handler/openai_jiutian_handler.go
new file mode 100644
index 0000000..86bbbb3
--- /dev/null
+++ b/pkg/handler/openai_jiutian_handler.go
@@ -0,0 +1,186 @@
+package handler
+
+import (
+	"bufio"
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"io"
+	"net/http"
+	"simple-one-api/pkg/adapter"
+	"simple-one-api/pkg/config"
+	"simple-one-api/pkg/llm/jiutian"
+	"simple-one-api/pkg/mylog"
+	"simple-one-api/pkg/utils"
+	"go.uber.org/zap"
+)
+
+// OpenAI2JiuTianHandler 处理OpenAI到九天模型的请求转换
+func OpenAI2JiuTianHandler(c *gin.Context, oaiReqParam *OAIRequestParam) error {
+	mylog.Logger.Info("Starting JiuTian request handling")
+	
+	oaiReq := oaiReqParam.chatCompletionReq
+	s := oaiReqParam.modelDetails
+
+	// 获取API Key
+	apiKey, _ := utils.GetStringFromMap(oaiReqParam.creds, config.KEYNAME_API_KEY)
+	if apiKey == "" {
+		return fmt.Errorf("API key not found")
+	}
+
+	// 转换请求
+	jiutianReq := adapter.OpenAIRequestToJiuTianRequest(oaiReq)
+	
+	// 分别设置各个参数
+	jiutianReq.WithAPIKey(apiKey)
+	jiutianReq.WithBaseURL(s.ServerURL)
+	
+	// 确保transport被正确设置
+	if oaiReqParam.httpTransport != nil {
+		mylog.Logger.Debug("Setting custom transport for JiuTian request")
+		jiutianReq.WithTransport(oaiReqParam.httpTransport)
+	} else {
+		mylog.Logger.Debug("Using default transport for JiuTian request")
+		jiutianReq.WithTransport(http.DefaultTransport)
+	}
+
+	// 处理流式请求
+	if oaiReq.Stream {
+		return handleJiuTianStreamRequest(c, jiutianReq, oaiReqParam.ClientModel)
+	}
+
+	// 处理非流式请求
+	return handleJiuTianNonStreamRequest(c, jiutianReq, oaiReqParam.ClientModel)
+}
+
+// handleJiuTianStreamRequest 处理流式请求
+func handleJiuTianStreamRequest(c *gin.Context, jiutianReq *jiutian.ChatCompletionRequest, clientModel string) error {
+	mylog.Logger.Info("Handling JiuTian stream request")
+
+	// 记录请求头信息
+	mylog.Logger.Info("Original request headers",
+		zap.Any("headers", c.Request.Header))
+
+	// 记录请求内容
+	reqData, _ := json.Marshal(jiutianReq)
+	mylog.Logger.Info("Request content",
+		zap.String("request_body", string(reqData)))
+
+	// 发送流式请求
+	resp, err := jiutianReq.CreateCompletionStream()
+	if err != nil {
+		return err
+	}
+	defer resp.Body.Close()
+
+	// 设置响应头
+	c.Header("Content-Type", "text/event-stream")
+	c.Header("Cache-Control", "no-cache")
+	c.Header("Connection", "keep-alive")
+
+	// 记录设置的响应头
+	mylog.Logger.Info("Set response headers",
+		zap.String("content_type", c.Writer.Header().Get("Content-Type")),
+		zap.String("cache_control", c.Writer.Header().Get("Cache-Control")),
+		zap.String("connection", c.Writer.Header().Get("Connection")))
+
+	// 读取并转发响应
+	reader := bufio.NewReader(resp.Body)
+	for {
+		line, err := reader.ReadBytes('\n')
+		if err != nil {
+			if err == io.EOF {
+				break
+			}
+			return err
+		}
+
+		// 处理SSE数据
+		if bytes.HasPrefix(line, []byte("data: ")) {
+			data := bytes.TrimPrefix(line, []byte("data: "))
+			// 去掉可能存在的换行符
+			data = bytes.TrimSpace(data)
+			
+			mylog.Logger.Info("Received stream data", 
+				zap.String("raw_data", string(data)),
+				zap.Int("data_length", len(data)))
+
+			var jiutianResp jiutian.ChatCompletionStreamResponse
+			if err := json.Unmarshal(data, &jiutianResp); err != nil {
+				mylog.Logger.Error("Failed to parse stream response",
+					zap.Error(err),
+					zap.String("data", string(data)))
+				continue
+			}
+
+			// 记录解析后的九天响应
+			mylog.Logger.Info("Parsed JiuTian response",
+				zap.Any("jiutian_response", map[string]interface{}{
+					"response": jiutianResp.Response,
+					"delta":    jiutianResp.Delta,
+					"finished": jiutianResp.Finished,
+					"history":  jiutianResp.History,
+				}))
+
+			// 转换为OpenAI流式响应
+			streamResp := adapter.JiuTianStreamResponseToOpenAIStreamResponse(&jiutianResp)
+			streamResp.Model = clientModel
+
+			// 发送响应
+			responseData, _ := json.Marshal(streamResp)
+			mylog.Logger.Info("Sending stream response",
+				zap.String("response_data", string(responseData)))
+
+			c.Writer.Write([]byte("data: "))
+			c.Writer.Write(responseData)
+			c.Writer.Write([]byte("\n\n"))
+			c.Writer.Flush()
+		}
+	}
+
+	return nil
+}
+
+// handleJiuTianNonStreamRequest 处理非流式请求
+func handleJiuTianNonStreamRequest(c *gin.Context, jiutianReq *jiutian.ChatCompletionRequest, clientModel string) error {
+	mylog.Logger.Info("Handling JiuTian non-stream request")
+
+	// 记录请求头信息
+	mylog.Logger.Info("Original request headers",
+		zap.Any("headers", c.Request.Header))
+
+	// 记录请求内容
+	reqData, _ := json.Marshal(jiutianReq)
+	mylog.Logger.Info("Request content",
+		zap.String("request_body", string(reqData)))
+
+	// 发送请求
+	jiutianResp, err := jiutianReq.CreateCompletion()
+	if err != nil {
+		return err
+	}
+
+	// 记录九天响应
+	mylog.Logger.Info("Received JiuTian response",
+		zap.Any("jiutian_response", map[string]interface{}{
+			"usage":    jiutianResp.Usage,
+			"response": jiutianResp.Response,
+			"delta":    jiutianResp.Delta,
+			"finished": jiutianResp.Finished,
+			"history":  jiutianResp.History,
+		}))
+
+	// 转换为OpenAI响应
+	chatResp := adapter.JiuTianResponseToOpenAIResponse(jiutianResp)
+	chatResp.Model = clientModel
+
+	// 记录响应信息
+	responseData, _ := json.Marshal(chatResp)
+	mylog.Logger.Info("Sending final response",
+		zap.String("response_data", string(responseData)))
+
+	// 发送响应
+	c.JSON(http.StatusOK, chatResp)
+	return nil
+} 
\ No newline at end of file
diff --git a/pkg/llm/jiutian/jiutian_request.go b/pkg/llm/jiutian/jiutian_request.go
new file mode 100644
index 0000000..df238cf
--- /dev/null
+++ b/pkg/llm/jiutian/jiutian_request.go
@@ -0,0 +1,396 @@
+package jiutian
+
+import (
+	"bytes"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"github.com/golang-jwt/jwt/v5"
+	"go.uber.org/zap"
+	"io"
+	"net/http"
+	"simple-one-api/pkg/mylog"
+	"strings"
+	"time"
+)
+
+const (
+	// DefaultBaseURL 默认的九天模型API基础URL
+	DefaultBaseURL = "https://jiutian.10086.cn/largemodel/api/v1"
+	// DefaultTimeout 默认超时时间
+	DefaultTimeout = 3 * time.Minute
+)
+
+// Message 九天模型的消息结构
+type Message struct {
+	Role    string `json:"role"`    // 角色：system/user/assistant
+	Content string `json:"content"` // 消息内容
+}
+
+// ChatCompletionRequest 九天模型的对话请求结构
+type ChatCompletionRequest struct {
+	ModelID     string     `json:"modelId"`              // 模型ID
+	Prompt      string     `json:"prompt"`               // 当前问题
+	Params      *Params    `json:"params"`               // 参数配置
+	History     [][]string `json:"history"`              // 历史对话记录
+	Stream      bool       `json:"stream"`               // 是否使用流式响应
+	apiKey      string     // API密钥
+	baseURL     string     // API基础URL
+	transport   http.RoundTripper // HTTP传输层
+}
+
+// Params 模型参数配置
+type Params struct {
+	Temperature float32 `json:"temperature"` // 温度参数
+	TopP        float32 `json:"top_p"`      // 核采样参数
+}
+
+// NewChatCompletionRequest 创建新的对话请求
+func NewChatCompletionRequest() *ChatCompletionRequest {
+	return &ChatCompletionRequest{
+		ModelID: "Llama3.1-70B", // 默认模型
+		Params: &Params{
+			Temperature: 0.7,  // 默认温度
+			TopP:        0.95, // 默认top_p
+		},
+		History: make([][]string, 0),
+		Stream:  false,
+		baseURL: DefaultBaseURL,
+	}
+}
+
+// WithModelID 设置模型ID
+func (r *ChatCompletionRequest) WithModelID(modelID string) *ChatCompletionRequest {
+	r.ModelID = modelID
+	return r
+}
+
+// WithPrompt 设置当前问题
+func (r *ChatCompletionRequest) WithPrompt(prompt string) *ChatCompletionRequest {
+	r.Prompt = prompt
+	return r
+}
+
+// WithHistory 设置历史对话记录
+func (r *ChatCompletionRequest) WithHistory(history [][]string) *ChatCompletionRequest {
+	r.History = history
+	return r
+}
+
+// WithStream 设置是否使用流式响应
+func (r *ChatCompletionRequest) WithStream(stream bool) *ChatCompletionRequest {
+	r.Stream = stream
+	return r
+}
+
+// WithTemperature 设置温度参数
+func (r *ChatCompletionRequest) WithTemperature(temperature float32) *ChatCompletionRequest {
+	r.Params.Temperature = temperature
+	return r
+}
+
+// WithTopP 设置top_p参数
+func (r *ChatCompletionRequest) WithTopP(topP float32) *ChatCompletionRequest {
+	r.Params.TopP = topP
+	return r
+}
+
+// WithAPIKey 设置API密钥
+func (r *ChatCompletionRequest) WithAPIKey(apiKey string) *ChatCompletionRequest {
+	r.apiKey = apiKey
+	return r
+}
+
+// WithBaseURL 设置API基础URL
+func (r *ChatCompletionRequest) WithBaseURL(baseURL string) *ChatCompletionRequest {
+	if baseURL != "" {
+		r.baseURL = baseURL
+	}
+	return r
+}
+
+// WithTransport 设置HTTP传输层
+func (r *ChatCompletionRequest) WithTransport(transport http.RoundTripper) *ChatCompletionRequest {
+	r.transport = transport
+	return r
+}
+
+// validate 验证请求参数
+func (r *ChatCompletionRequest) validate() error {
+	if r.ModelID == "" {
+		return errors.New("modelId is required")
+	}
+	if r.Prompt == "" {
+		return errors.New("prompt is required")
+	}
+	if r.Params == nil {
+		return errors.New("params is required")
+	}
+	if r.Params.Temperature < 0 || r.Params.Temperature > 2 {
+		return errors.New("temperature must be between 0 and 2")
+	}
+	if r.Params.TopP < 0 || r.Params.TopP > 1 {
+		return errors.New("top_p must be between 0 and 1")
+	}
+	if r.apiKey == "" {
+		return errors.New("API key is required")
+	}
+	return nil
+}
+
+// generateToken 生成JWT token
+func (r *ChatCompletionRequest) generateToken() (string, error) {
+	mylog.Logger.Info("Generating JiuTian JWT token")
+
+	// 分割API Key
+	parts := strings.Split(r.apiKey, ".")
+	if len(parts) != 2 {
+		mylog.Logger.Error("Invalid API key format", zap.String("api_key", r.apiKey))
+		return "", errors.New("invalid API key format")
+	}
+	id, secret := parts[0], parts[1]
+	mylog.Logger.Debug("API key parsed", zap.String("id", id))
+
+	// 创建token
+	now := time.Now().Unix()
+	claims := jwt.MapClaims{
+		"api_key":   id,
+		"exp":       now + 3600, // 1小时有效期
+		"timestamp": now,
+	}
+
+	mylog.Logger.Debug("Creating JWT claims",
+		zap.String("api_key", id),
+		zap.Int64("exp", now+3600),
+		zap.Int64("timestamp", now))
+
+	// 设置header
+	token := jwt.NewWithClaims(jwt.SigningMethodHS256, claims)
+	token.Header["alg"] = "HS256"
+	token.Header["typ"] = "JWT"
+	token.Header["sign_type"] = "SIGN"
+
+	mylog.Logger.Debug("JWT header set",
+		zap.String("alg", "HS256"),
+		zap.String("typ", "JWT"),
+		zap.String("sign_type", "SIGN"))
+
+	// 签名
+	tokenString, err := token.SignedString([]byte(secret))
+	if err != nil {
+		mylog.Logger.Error("Failed to sign token",
+			zap.Error(err),
+			zap.String("id", id))
+		return "", fmt.Errorf("failed to sign token: %v", err)
+	}
+
+	mylog.Logger.Info("JWT token generated successfully",
+		zap.String("token", tokenString),
+		zap.Int64("expires_at", now+3600))
+
+	return tokenString, nil
+}
+
+// CreateCompletion 发送非流式请求
+func (r *ChatCompletionRequest) CreateCompletion() (*ChatCompletionResponse, error) {
+	mylog.Logger.Info("Creating JiuTian chat completion")
+
+	// 验证请求参数
+	if err := r.validate(); err != nil {
+		return nil, err
+	}
+
+	// 生成token
+	token, err := r.generateToken()
+	if err != nil {
+		mylog.Logger.Error("Failed to generate token", zap.Error(err))
+		return nil, err
+	}
+
+	// 准备请求URL
+	url := fmt.Sprintf("%s/completions", r.baseURL)
+	mylog.Logger.Info("Accessing JiuTian API", zap.String("full_url", url))
+
+	// 准备请求数据
+	jsonData, err := json.Marshal(r)
+	if err != nil {
+		return nil, err
+	}
+
+	// 创建HTTP客户端
+	client := &http.Client{
+		Timeout: DefaultTimeout,
+	}
+	
+	// 设置Transport
+	if r.transport != nil {
+		client.Transport = r.transport
+	} else {
+		client.Transport = http.DefaultTransport
+	}
+
+	// 创建HTTP请求
+	httpReq, err := http.NewRequest("POST", url, bytes.NewBuffer(jsonData))
+	if err != nil {
+		return nil, err
+	}
+
+	// 设置请求头
+	httpReq.Header.Set("Content-Type", "application/json")
+	httpReq.Header.Set("Authorization", fmt.Sprintf("Bearer %s", token))
+
+	// 记录完整的请求信息
+	mylog.Logger.Info("Request details",
+		zap.String("url", url),
+		zap.String("method", httpReq.Method),
+		zap.String("content_type", httpReq.Header.Get("Content-Type")),
+		zap.String("authorization", "Bearer "+token[:10]+"..."), // 只显示token的前10个字符
+		zap.String("request_body", string(jsonData)))
+
+	// 发送请求
+	resp, err := client.Do(httpReq)
+	if err != nil {
+		mylog.Logger.Error("Failed to send request",
+			zap.Error(err),
+			zap.String("url", url))
+		return nil, err
+	}
+	defer resp.Body.Close()
+
+	// 记录响应头信息
+	mylog.Logger.Info("Response headers",
+		zap.Int("status_code", resp.StatusCode),
+		zap.String("content_type", resp.Header.Get("Content-Type")),
+		zap.String("content_length", resp.Header.Get("Content-Length")),
+		zap.Any("all_headers", resp.Header))
+
+	// 读取响应
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
+	// 检查响应状态码
+	if resp.StatusCode != http.StatusOK {
+		mylog.Logger.Error("API request failed",
+			zap.Int("status_code", resp.StatusCode),
+			zap.String("body", string(body)))
+		return nil, fmt.Errorf("API request failed with status code: %d, body: %s", resp.StatusCode, string(body))
+	}
+
+	// 处理响应数据
+	responseStr := string(body)
+	mylog.Logger.Debug("Raw response", zap.String("body", responseStr))
+
+	// 如果响应以 "data:" 开头，需要去掉这个前缀
+	if strings.HasPrefix(responseStr, "data:") {
+		responseStr = strings.TrimPrefix(responseStr, "data:")
+		// 去掉可能存在的换行符
+		responseStr = strings.TrimSpace(responseStr)
+	}
+
+	// 解析响应
+	var response ChatCompletionResponse
+	if err := json.Unmarshal([]byte(responseStr), &response); err != nil {
+		mylog.Logger.Error("Failed to parse response",
+			zap.Error(err),
+			zap.String("body", responseStr))
+		return nil, err
+	}
+
+	mylog.Logger.Debug("Received response from JiuTian API",
+		zap.Any("usage", response.Usage),
+		zap.String("response", response.Response),
+		zap.String("finished", response.Finished))
+
+	return &response, nil
+}
+
+// CreateCompletionStream 发送流式请求
+func (r *ChatCompletionRequest) CreateCompletionStream() (*http.Response, error) {
+	mylog.Logger.Info("Creating JiuTian stream chat completion")
+
+	// 验证请求参数
+	if err := r.validate(); err != nil {
+		return nil, err
+	}
+
+	// 生成token
+	token, err := r.generateToken()
+	if err != nil {
+		mylog.Logger.Error("Failed to generate token", zap.Error(err))
+		return nil, err
+	}
+
+	// 准备请求URL
+	url := fmt.Sprintf("%s/completions", r.baseURL)
+	mylog.Logger.Info("Accessing JiuTian API Stream", zap.String("full_url", url))
+
+	// 准备请求数据
+	jsonData, err := json.Marshal(r)
+	if err != nil {
+		return nil, err
+	}
+
+	// 创建HTTP客户端
+	client := &http.Client{
+		Timeout: DefaultTimeout,
+	}
+	
+	// 设置Transport
+	if r.transport != nil {
+		client.Transport = r.transport
+	} else {
+		client.Transport = http.DefaultTransport
+	}
+
+	// 创建HTTP请求
+	httpReq, err := http.NewRequest("POST", url, bytes.NewBuffer(jsonData))
+	if err != nil {
+		return nil, err
+	}
+
+	// 设置请求头
+	httpReq.Header.Set("Content-Type", "application/json")
+	httpReq.Header.Set("Authorization", fmt.Sprintf("Bearer %s", token))
+	httpReq.Header.Set("Accept", "text/event-stream")
+	httpReq.Header.Set("Cache-Control", "no-cache")
+	httpReq.Header.Set("Connection", "keep-alive")
+
+	// 记录完整的请求信息
+	mylog.Logger.Info("Stream request details",
+		zap.String("url", url),
+		zap.String("method", httpReq.Method),
+		zap.String("content_type", httpReq.Header.Get("Content-Type")),
+		zap.String("accept", httpReq.Header.Get("Accept")),
+		zap.String("cache_control", httpReq.Header.Get("Cache-Control")),
+		zap.String("connection", httpReq.Header.Get("Connection")),
+		zap.String("authorization", "Bearer "+token[:10]+"..."), // 只显示token的前10个字符
+		zap.String("request_body", string(jsonData)))
+
+	// 发送请求
+	resp, err := client.Do(httpReq)
+	if err != nil {
+		mylog.Logger.Error("Failed to send stream request",
+			zap.Error(err),
+			zap.String("url", url))
+		return nil, err
+	}
+
+	// 记录响应头信息
+	mylog.Logger.Info("Stream response headers",
+		zap.Int("status_code", resp.StatusCode),
+		zap.String("content_type", resp.Header.Get("Content-Type")),
+		zap.String("transfer_encoding", resp.Header.Get("Transfer-Encoding")),
+		zap.Any("all_headers", resp.Header))
+
+	// 检查响应状态码
+	if resp.StatusCode != http.StatusOK {
+		resp.Body.Close()
+		mylog.Logger.Error("Stream API request failed",
+			zap.Int("status_code", resp.StatusCode))
+		return nil, fmt.Errorf("API request failed with status code: %d", resp.StatusCode)
+	}
+
+	return resp, nil
+} 
\ No newline at end of file
diff --git a/pkg/llm/jiutian/jiutian_response.go b/pkg/llm/jiutian/jiutian_response.go
new file mode 100644
index 0000000..cb2f4db
--- /dev/null
+++ b/pkg/llm/jiutian/jiutian_response.go
@@ -0,0 +1,25 @@
+package jiutian
+
+// Usage 使用统计
+type Usage struct {
+	PromptTokens     int `json:"prompt_tokens"`     // 输入的token数量
+	CompletionTokens int `json:"completion_tokens"` // 生成的token数量
+	TotalTokens      int `json:"total_tokens"`      // 总token数量
+}
+
+// ChatCompletionResponse 九天模型的对话响应结构
+type ChatCompletionResponse struct {
+	Usage    Usage     `json:"Usage"`    // 使用统计
+	Response string    `json:"response"` // 模型回答
+	Delta    string    `json:"delta"`    // 结束标记
+	Finished string    `json:"finished"` // 结束原因
+	History  [][]string `json:"history"` // 历史对话记录
+}
+
+// ChatCompletionStreamResponse 九天模型的流式响应结构
+type ChatCompletionStreamResponse struct {
+	Response string    `json:"response"` // 当前生成的内容
+	Delta    string    `json:"delta"`    // 结束标记
+	Finished string    `json:"finished"` // 结束原因
+	History  [][]string `json:"history"` // 历史对话记录
+} 
\ No newline at end of file