diff --git a/README.md b/README.md
index fe91dbc7..1dddc4b0 100644
--- a/README.md
+++ b/README.md
@@ -68,12 +68,17 @@ Get up and running with WhoDB quickly using Docker:
 docker run -it -p 8080:8080 clidey/whodb
 ```
 
+
+To run WhoDB with an OpenAI compatible service, you should assign some environments.
+```sh
+docker run -it -e USE_CUSTOM_MODELS=1 -e CUSTOM_MODELS=gpt-4o,gpt-3.5,others -e OPENAI_BASE_URL=http://your_base_url/v1 -p 8080:8080 clidey/whodb
+```
+
 If you are using a remote Ollama server, please start Docker with Ollama Environments like this:
 ```sh
 docker run -it -e WHODB_OLLAMA_HOST=YOUR_OLLAMA_HOST -e WHODB_OLLAMA_PORT=YOUR_OLLAMA_PORT -p 8080:8080 clidey/whodb
 ```
 
-
 Or, use Docker Compose:
 ```sh
 version: "3.8"
diff --git a/core/src/llm/chatgpt_client.go b/core/src/llm/chatgpt_client.go
index 6996804f..77fdab31 100644
--- a/core/src/llm/chatgpt_client.go
+++ b/core/src/llm/chatgpt_client.go
@@ -9,7 +9,6 @@ import (
 	"strings"
 )
 
-const chatGPTEndpoint = "https://api.openai.com/v1"
 
 func prepareChatGPTRequest(c *LLMClient, prompt string, model LLMModel, receiverChan *chan string) (string, []byte, map[string]string, error) {
 	requestBody, err := json.Marshal(map[string]interface{}{
@@ -20,7 +19,7 @@ func prepareChatGPTRequest(c *LLMClient, prompt string, model LLMModel, receiver
 	if err != nil {
 		return "", nil, nil, err
 	}
-	url := fmt.Sprintf("%v/chat/completions", chatGPTEndpoint)
+	url := fmt.Sprintf("%v/chat/completions", getOpenAICompatibleBaseURL())
 	headers := map[string]string{
 		"Authorization": fmt.Sprintf("Bearer %s", c.APIKey),
 		"Content-Type":  "application/json",
@@ -29,7 +28,7 @@ func prepareChatGPTRequest(c *LLMClient, prompt string, model LLMModel, receiver
 }
 
 func prepareChatGPTModelsRequest(apiKey string) (string, map[string]string) {
-	url := fmt.Sprintf("%v/models", chatGPTEndpoint)
+	url := fmt.Sprintf("%v/models", getOpenAICompatibleBaseURL())
 	headers := map[string]string{
 		"Authorization": fmt.Sprintf("Bearer %s", apiKey),
 		"Content-Type":  "application/json",
diff --git a/core/src/llm/env.go b/core/src/llm/env.go
index 6718d2a7..2bb7bb6f 100644
--- a/core/src/llm/env.go
+++ b/core/src/llm/env.go
@@ -2,7 +2,8 @@ package llm
 
 import (
 	"fmt"
-
+    "strings"
+    "os"
 	"github.com/clidey/whodb/core/src/common"
 	"github.com/clidey/whodb/core/src/env"
 )
@@ -24,3 +25,31 @@ func getOllamaEndpoint() string {
 
 	return fmt.Sprintf("http://%v:%v/api", host, port)
 }
+
+func getOpenAICompatibleBaseURL() string {
+	defaultBaseURL := "https://api.openai.com/v1"
+	baseURL := os.Getenv("OPENAI_BASE_URL")
+	if baseURL == "" {
+		baseURL = defaultBaseURL
+	}
+	return baseURL
+}
+
+func getCustomModels() ([]string, error) {
+    modelsStr := os.Getenv("CUSTOM_MODELS")
+	if modelsStr == "" {
+		return []string{}, nil
+	}
+
+	models := strings.Split(modelsStr, ",")
+
+	for i := range models {
+		models[i] = strings.TrimSpace(models[i])
+	}
+    return models, nil
+}
+
+func ShouldUseCustomModels() bool {
+	useCustomModels := os.Getenv("USE_CUSTOM_MODELS")
+	return useCustomModels == "1"
+}
\ No newline at end of file
diff --git a/core/src/llm/llm_client.go b/core/src/llm/llm_client.go
index 709a04ec..557cab27 100644
--- a/core/src/llm/llm_client.go
+++ b/core/src/llm/llm_client.go
@@ -66,6 +66,9 @@ func (c *LLMClient) GetSupportedModels() ([]string, error) {
 		url, headers = prepareOllamaModelsRequest()
 	case ChatGPT_LLMType:
 		url, headers = prepareChatGPTModelsRequest(c.APIKey)
+		if ShouldUseCustomModels() {
+		    return getCustomModels()
+		}
 	case Anthropic_LLMType:
 		return getAnthropicModels(c.APIKey)
 	default: