alibaba/spring-ai-alibaba · error · RuntimeException

模型配置不存在:

Error message

模型配置不存在: 

What it means

ChatClientFactoryDelegate.createChatClient first looks up the model config in a bridging service, then falls back to modelConfigRepository.findById. If both return null, it throws this RuntimeException because a ChatClient cannot be built without a model configuration record for the given id.

Source

Thrown at spring-ai-alibaba-admin/spring-ai-alibaba-admin-server-start/src/main/java/com/alibaba/cloud/ai/studio/admin/service/client/ChatClientFactoryDelegate.java:85

        return createChatClient(modelConfigId, userParameters, null,null);
    }
    
    public ChatClient createChatClient(Long modelConfigId, Map<String, Object> userParameters, Map<String, String> observationMetadata){
        return createChatClient(modelConfigId, userParameters, null, observationMetadata);
    }
    
    public ChatClient createChatClient(Long modelConfigId, Map<String, Object> userParameters, List<Advisor> advisors, Map<String, String> observationMetadata) {
        // 优先从桥接服务查找（从Manager层查询）
        ModelConfigDO config = modelConfigBridgeService.findById(modelConfigId);
        
        // 如果桥接服务中找不到，尝试从ModelConfigRepository查找（保持向后兼容）
        if (config == null) {
            log.debug("桥接服务中未找到模型配置 id={}，尝试从 ModelConfigRepository 查找", modelConfigId);
            config = modelConfigRepository.findById(modelConfigId);
        }
        
        if (config == null) {
            throw new RuntimeException("模型配置不存在: " + modelConfigId);
        }
        
        if (config.getStatus() != 1) {
            throw new RuntimeException("模型配置已禁用: " + modelConfigId);
        }

        String provider = config.getProvider().toLowerCase();
        log.info("创建模型客户端，提供商: {}, 模型: {}", provider, config.getModelName());
        
        ChatClientFactory factory = chatClientFactories.get(provider);
        if (factory == null) {
            // 如果找不到对应的 provider factory，默认使用 OpenAI
            log.warn("未找到提供商 {} 对应的工厂，使用默认的 OpenAI 工厂", provider);
            factory = chatClientFactories.get(OpenAiChatClientFactory.OPEN_AI_PROVIDER);
            if (factory == null) {
                throw new UnsupportedOperationException("不支持的模型提供商: " + config.getProvider() + "，且默认的 OpenAI 工厂也不可用");
            }
        }

View on GitHub (pinned to f82da0b50f)

Solutions

1. Verify the modelConfigId exists: check the models YAML file or the model config store for an entry with that id
2. Re-create or re-point the session/request to a valid existing model config id
3. Check startup logs for FileModelConfigRepository load failures (apiKey/duplicate id/name validation) that would leave the repository empty
4. If configs are managed in the admin UI, confirm the model config was not deleted or disabled by another user

Example fix

// before
chatClientFactoryDelegate.createChatClient(999L, params, metadata); // id never existed
// after
Long id = modelConfigRepository.findByName("qwen-max").getId();
chatClientFactoryDelegate.createChatClient(id, params, metadata);

// Java: check existence before creating the client
if (modelConfigRepository.findById(modelConfigId) == null) {
    throw new IllegalStateException("Resolve a valid modelConfigId first: " + modelConfigId);
}

boolean configExists(ModelConfigDO c) { return c != null; }

try {
    client = delegate.createChatClient(modelConfigId, params, metadata);
} catch (RuntimeException e) {
    if (e.getMessage().startsWith("模型配置不存在")) {
        // re-resolve or recreate the model config before retrying
    }
}

When it happens

Trigger: Calling createChatClient(modelConfigId, ...) with an id that exists neither in the bridge service nor the repository — e.g., a stale modelConfigId from a deleted session, or a config file that failed to load/validate (see errors 440-442).

Common situations: Sessions persisted before the model config was removed; YAML model file edited/renumbered so old ids no longer exist; repository failed to load due to a validation error earlier; case/format mismatch in the id passed by the caller.

Understand the failure class

Background: "Not found" and "does not exist" errors: why "Task not found", "No such folder", and "Can't find" fire when a lookup comes back empty — this error's family across 14 libraries.

Related errors

• 会话不存在:
• 数据集版本不存在:
• 模型配置不存在: modelId=
• AppNotFound
• AppConfigNotFound

AI-assisted analysis of alibaba/spring-ai-alibaba@f82da0b50f (2026-09-09). Data as JSON: /api/errors/8242e23489c61845. Report an issue: GitHub.