feat(api): DeepSeek V4 Flash 默认、HTTP 错讯与多供应商分层

- 主链路默认 deepseek-v4-flash，DEEPSEEK_THINKING_ENABLED 对齐旧非思考 chat
- 评测台评审装配迁入 adapters/llm（deepseek_eval_judge、zhipu_eval_judge）与 eval_judge_spec
- 拆分 llm_http_openai_chat_errors 与 llm_errors（DeepSeek/智谱品牌与文档链），llm_call 支持 http_error_vendor
- EvalJudgeService 按 spec.provider 传入 allm_json_call；评测台前端文案改为 V4 Flash
- 更新 .env 示例与 staging/production 的 DEEPSEEK_MODEL；补充 openai/供应商错讯测试

Made-with: Cursor

api/app/adapters/llm/deepseek.py

@@ -18,8 +18,10 @@ class DeepSeekLLMProvider:
         self,
         api_key: str,
         base_url: str = "https://api.deepseek.com",
-        model: str = "deepseek-chat",
+        model: str = "deepseek-v4-flash",
         temperature: float = 0.7,
+        *,
+        extra_body: dict | None = None,
     ):
         self._default_model = model
         self._default_temperature = temperature
@@ -28,6 +30,8 @@ class DeepSeekLLMProvider:
             "model": model,
             "api_key": api_key,
         }
+        if extra_body:
+            kwargs["extra_body"] = extra_body
         if base_url:
             cleaned = base_url.rstrip("/")
             for suffix in ("/v1/chat/completions", "/v1"):