DeepSeek Chat

Spring AI supports the various AI language models from DeepSeek. You can interact with DeepSeek language models and create a multilingual conversational assistant based on DeepSeek models.

Prerequisites

You will need to create an API key with DeepSeek to access DeepSeek language models. Create an account at DeepSeek registration page and generate a token on the API Keys page. The Spring AI project defines a configuration property named spring.ai.deepseek.api-key that you should set to the value of the API Key obtained from the API Keys page. Exporting an environment variable is one way to set this configuration property:

export SPRING_AI_DEEPSEEK_API_KEY=<INSERT KEY HERE>

Add Repositories and BOM

Spring AI artifacts are published in the Spring Milestone and Snapshot repositories. Refer to the Repositories section to add these repositories to your build system.

To help with dependency management, Spring AI provides a BOM (bill of materials) to ensure that a consistent version of Spring AI is used throughout your entire project. Refer to the Dependency Management section to add the Spring AI BOM to your build system.

Auto-configuration

Spring AI provides Spring Boot auto-configuration for the DeepSeek Chat Model. To enable it, add the following dependency to your project’s Maven pom.xml file:

<dependency>
    <groupId>org.springframework.ai</groupId>
    <artifactId>spring-ai-deepseek-spring-boot-starter</artifactId>
</dependency>

or to your Gradle build.gradle file.

dependencies {
    implementation 'org.springframework.ai:spring-ai-deepseek-spring-boot-starter'
}
Refer to the Dependency Management section to add the Spring AI BOM to your build file.

Chat Properties

Retry Properties

The prefix spring.ai.retry is used as the property prefix that lets you configure the retry mechanism for the DeepSeek Chat model.

Property Description Default

spring.ai.retry.max-attempts

Maximum number of retry attempts.

10

spring.ai.retry.backoff.initial-interval

Initial sleep duration for the exponential backoff policy.

2 sec.

spring.ai.retry.backoff.multiplier

Backoff interval multiplier.

5

spring.ai.retry.backoff.max-interval

Maximum backoff duration.

3 min.

spring.ai.retry.on-client-errors

If false, throws a NonTransientAiException, and does not attempt a retry for 4xx client error codes

false

spring.ai.retry.exclude-on-http-codes

List of HTTP status codes that should not trigger a retry (e.g. to throw NonTransientAiException).

empty

spring.ai.retry.on-http-codes

List of HTTP status codes that should trigger a retry (e.g. to throw TransientAiException).

empty

Connection Properties

The prefix spring.ai.deepseek is used as the property prefix that lets you connect to DeepSeek.

Property Description Default

spring.ai.deepseek.base-url

The URL to connect to

api.deepseek.com

spring.ai.deepseek.api-key

The API Key

-

Configuration Properties

The prefix spring.ai.deepseek.chat is the property prefix that lets you configure the chat model implementation for DeepSeek.

Property Description Default

spring.ai.deepseek.chat.enabled

Enables the DeepSeek chat model.

true

spring.ai.deepseek.chat.base-url

Optionally overrides the spring.ai.deepseek.base-url to provide a chat-specific URL

api.deepseek.com/

spring.ai.deepseek.chat.api-key

Optionally overrides the spring.ai.deepseek.api-key to provide a chat-specific API key

-

spring.ai.deepseek.chat.completions-path

The path to the chat completions endpoint

/chat/completions

spring.ai.deepseek.chat.beta-prefix-path

The prefix path to the beta feature endpoint

/beta/chat/completions

spring.ai.deepseek.chat.options.model

ID of the model to use. You can use either deepseek-coder or deepseek-chat.

deepseek-chat

spring.ai.deepseek.chat.options.frequencyPenalty

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model’s likelihood to repeat the same line verbatim.

0.0f

spring.ai.deepseek.chat.options.maxTokens

The maximum number of tokens to generate in the chat completion. The total length of input tokens and generated tokens is limited by the model’s context length.

-

spring.ai.deepseek.chat.options.presencePenalty

Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model’s likelihood to talk about new topics.

0.0f

spring.ai.deepseek.chat.options.stop

Up to 4 sequences where the API will stop generating further tokens.

-

spring.ai.deepseek.chat.options.temperature

Which sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or top_p, but not both.

1.0F

spring.ai.deepseek.chat.options.topP

An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or temperature, but not both.

1.0F

spring.ai.deepseek.chat.options.logprobs

Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of the message.

-

spring.ai.deepseek.chat.options.topLogprobs

An integer between 0 and 20 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to true if this parameter is used.

-
You can override the common spring.ai.deepseek.base-url and spring.ai.deepseek.api-key for the ChatModel implementations. The spring.ai.deepseek.chat.base-url and spring.ai.deepseek.chat.api-key properties, if set, take precedence over the common properties. This is useful if you want to use different DeepSeek accounts for different models and different model endpoints.
All properties prefixed with spring.ai.deepseek.chat.options can be overridden at runtime by adding a request-specific Runtime Options to the Prompt call.

Runtime Options

The DeepSeekChatOptions.java provides model configurations, such as the model to use, the temperature, the frequency penalty, etc.

On startup, the default options can be configured with the DeepSeekChatModel(api, options) constructor or the spring.ai.deepseek.chat.options.* properties.

At runtime, you can override the default options by adding new, request-specific options to the Prompt call. For example, to override the default model and temperature for a specific request:

ChatResponse response = chatModel.call(
    new Prompt(
        "Generate the names of 5 famous pirates. Please provide the JSON response without any code block markers such as ```json```.",
        DeepSeekChatOptions.builder()
            .withModel(DeepSeekApi.ChatModel.DEEPSEEK_CHAT.getValue())
            .withTemperature(0.8f)
        .build()
    ));
In addition to the model-specific DeepSeekChatOptions, you can use a portable ChatOptions instance, created with the ChatOptionsBuilder#builder().

Sample Controller (Auto-configuration)

Create a new Spring Boot project and add the spring-ai-deepseek-spring-boot-starter to your pom (or gradle) dependencies.

Add an application.properties file under the src/main/resources directory to enable and configure the DeepSeek Chat model:

spring.ai.deepseek.api-key=YOUR_API_KEY
spring.ai.deepseek.chat.options.model=deepseek-chat
spring.ai.deepseek.chat.options.temperature=0.8
Replace the api-key with your DeepSeek credentials.

This will create a DeepSeekChatModel implementation that you can inject into your class. Here is an example of a simple @Controller class that uses the chat model for text generation.

@RestController
public class ChatController {

    private final DeepSeekChatModel chatModel;

    @Autowired
    public ChatController(DeepSeekChatModel chatModel) {
        this.chatModel = chatModel;
    }

    @GetMapping("/ai/generate")
    public Map generate(@RequestParam(value = "message", defaultValue = "Tell me a joke") String message) {
        return Map.of("generation", chatModel.call(message));
    }

    @GetMapping("/ai/generateStream")
	public Flux<ChatResponse> generateStream(@RequestParam(value = "message", defaultValue = "Tell me a joke") String message) {
        var prompt = new Prompt(new UserMessage(message));
        return chatModel.stream(prompt);
    }
}

Chat Prefix Completion

The chat prefix completion follows the Chat Completion API, where users provide an assistant’s prefix message for the model to complete the rest of the message.

When using prefix completion, the user must ensure that the last message in the messages list is a DeepSeekAssistantMessage.

Below is a complete Java code example for chat prefix completion. In this example, we set the prefix message of the assistant to "```python\n" to force the model to output Python code, and set the stop parameter to [‘`’] to prevent additional explanations from the model.

@RestController
public class CodeGenerateController {

    private final DeepSeekChatModel chatModel;

    @Autowired
    public ChatController(DeepSeekChatModel chatModel) {
        this.chatModel = chatModel;
    }

    @GetMapping("/ai/generatePythonCode")
    public String generate(@RequestParam(value = "message", defaultValue = "Please write quick sort code") String message) {
		UserMessage userMessage = new UserMessage(message);
		Message assistantMessage = DeepSeekAssistantMessage.prefixAssistantMessage("```python\\n");
		Prompt prompt = new Prompt(List.of(userMessage, assistantMessage), ChatOptions.builder().stopSequences(List.of("```")).build());
		ChatResponse response = chatModel.call(prompt);
		return response.getResult().getOutput().getText();
    }
}

Reasoning Model (deepseek-reasoner)

The deepseek-reasoner is a reasoning model developed by DeepSeek. Before delivering the final answer, the model first generates a Chain of Thought (CoT) to enhance the accuracy of its responses. Our API provides users with access to the CoT content generated by deepseek-reasoner, enabling them to view, display, and distill it.

You can use the DeepSeekAssistantMessage to get the CoT content generated by deepseek-reasoner.

public void deepSeekReasonerExample() {
    DeepSeekChatOptions promptOptions = DeepSeekChatOptions.builder()
            .model(DeepSeekApi.ChatModel.DEEPSEEK_REASONER.getValue())
            .build();
    Prompt prompt = new Prompt("9.11 and 9.8, which is greater?", promptOptions);
    ChatResponse response = chatModel.call(prompt);

    // Get the CoT content generated by deepseek-reasoner, only available when using deepseek-reasoner model
    DeepSeekAssistantMessage deepSeekAssistantMessage = (DeepSeekAssistantMessage) response.getResult().getOutput();
    String reasoningContent = deepSeekAssistantMessage.getReasoningContent();
    String text = deepSeekAssistantMessage.getText();
}

Reasoning Model Multi-round Conversation

In each round of the conversation, the model outputs the CoT (reasoning_content) and the final answer (content). In the next round of the conversation, the CoT from previous rounds is not concatenated into the context, as illustrated in the following diagram:

Multimodal Test Image

Please note that if the reasoning_content field is included in the sequence of input messages, the API will return a 400 error. Therefore, you should remove the reasoning_content field from the API response before making the API request, as demonstrated in the API example.

public String deepSeekReasonerMultiRoundExample() {
    List<Message> messages = new ArrayList<>();
    messages.add(new UserMessage("9.11 and 9.8, which is greater?"));
    DeepSeekChatOptions promptOptions = DeepSeekChatOptions.builder()
            .model(DeepSeekApi.ChatModel.DEEPSEEK_REASONER.getValue())
            .build();

    Prompt prompt = new Prompt(messages, promptOptions);
    ChatResponse response = chatModel.call(prompt);

    DeepSeekAssistantMessage deepSeekAssistantMessage = (DeepSeekAssistantMessage) response.getResult().getOutput();
    String reasoningContent = deepSeekAssistantMessage.getReasoningContent();
    String text = deepSeekAssistantMessage.getText();

    messages.add(new AssistantMessage(Objects.requireNonNull(text)));
    messages.add(new UserMessage("How many Rs are there in the word 'strawberry'?"));
    Prompt prompt2 = new Prompt(messages, promptOptions);
    ChatResponse response2 = chatModel.call(prompt2);

    DeepSeekAssistantMessage deepSeekAssistantMessage2 = (DeepSeekAssistantMessage) response2.getResult().getOutput();
    String reasoningContent2 = deepSeekAssistantMessage2.getReasoningContent();
    return deepSeekAssistantMessage2.getText();
}

Manual Configuration

The DeepSeekChatModel implements the ChatModel and StreamingChatModel and uses the Low-level DeepSeekApi Client to connect to the DeepSeek service.

Add the spring-ai-deepseek dependency to your project’s Maven pom.xml file:

<dependency>
    <groupId>org.springframework.ai</groupId>
    <artifactId>spring-ai-deepseek</artifactId>
</dependency>

or to your Gradle build.gradle file.

dependencies {
    implementation 'org.springframework.ai:spring-ai-deepseek'
}
Refer to the Dependency Management section to add the Spring AI BOM to your build file.

Next, create a DeepSeekChatModel and use it for text generation:

var deepSeekApi = new DeepSeekApi(System.getenv("DEEPSEEK_API_KEY"));

var chatModel = new DeepSeekChatModel(deepSeekApi, DeepSeekChatOptions.builder()
                .withModel(DeepSeekApi.ChatModel.DEEPSEEK_CHAT.getValue())
                .withTemperature(0.4f)
                .withMaxTokens(200)
                .build());

ChatResponse response = chatModel.call(
    new Prompt("Generate the names of 5 famous pirates."));

// Or with streaming responses
Flux<ChatResponse> streamResponse = chatModel.stream(
    new Prompt("Generate the names of 5 famous pirates."));

The DeepSeekChatOptions provides the configuration information for the chat requests. The DeepSeekChatOptions.Builder is a fluent options builder.

Low-level DeepSeekApi Client

The DeepSeekApi is a lightweight Java client for DeepSeek API.

Here is a simple snippet showing how to use the API programmatically:

DeepSeekApi deepSeekApi =
    new DeepSeekApi(System.getenv("DEEPSEEK_API_KEY"));

ChatCompletionMessage chatCompletionMessage =
    new ChatCompletionMessage("Hello world", Role.USER);

// Sync request
ResponseEntity<ChatCompletion> response = deepSeekApi.chatCompletionEntity(
    new ChatCompletionRequest(List.of(chatCompletionMessage), DeepSeekApi.ChatModel.DEEPSEEK_CHAT.getValue(), 0.7f, false));

// Streaming request
Flux<ChatCompletionChunk> streamResponse = deepSeekApi.chatCompletionStream(
        new ChatCompletionRequest(List.of(chatCompletionMessage), DeepSeekApi.ChatModel.DEEPSEEK_CHAT.getValue(), 0.7f, true));

Follow the DeepSeekApi.java's JavaDoc for further information.

DeepSeekApi Samples

  • The DeepSeekApiIT.java test provides some general examples of how to use the lightweight library.

OSZAR »