diff --git a/datasource/models.py b/datasource/models.py
new file mode 100644
index 0000000..5f89e86
--- /dev/null
+++ b/datasource/models.py
@@ -0,0 +1,359 @@
+"""
+Pydantic 数据模型
+==================
+定义 Flask API 的请求和响应数据结构
+
+用途:
+- 类型安全的 API 响应结构
+- 调用方数据验证
+- IDE 自动补全支持
+- 为 Phase 2 (Flask-Pydantic 集成) 做准备
+
+使用示例:
+    from datasource.models import OHLCVResponse
+    
+    # 验证 API 响应
+    response = requests.get(url).json()
+    validated = OHLCVResponse.model_validate(response)
+    
+    # 访问字段（有类型提示）
+    print(validated.code)
+    print(validated.latest_premium)
+"""
+
+from pydantic import BaseModel, Field, field_validator
+from typing import Optional, List, Dict, Any
+from enum import Enum
+
+
+# ============================================================
+# 枚举类型
+# ============================================================
+
+class AssetTypeEnum(str, Enum):
+    """资产类型枚举"""
+    CHINA_INDEX = "china_index"
+    CHINA_ETF = "china_etf"
+    CHINA_STOCK = "china_stock"
+    US_INDEX = "us_index"
+    US_STOCK = "us_stock"
+    HK_INDEX = "hk_index"
+    HK_STOCK = "hk_stock"
+    FUTURES = "futures"
+    CRYPTO = "crypto"
+
+
+class AdjTypeEnum(str, Enum):
+    """复权类型枚举"""
+    RAW = "raw"
+    QFQ = "qfq"
+    HFQ = "hfq"
+
+
+class TimeframeEnum(str, Enum):
+    """K线周期枚举（加密货币）"""
+    D1 = "1d"
+    H1 = "1h"
+    H4 = "4h"
+    M15 = "15m"
+    M1 = "1m"
+
+
+# ============================================================
+# 请求模型（Phase 2 使用）
+# ============================================================
+
+class OHLCVRequest(BaseModel):
+    """
+    OHLCV 数据获取请求
+    
+    使用示例:
+        request = OHLCVRequest(
+            code="META",
+            start="2024-01-01",
+            end="2024-03-31",
+            adj=AdjTypeEnum.RAW
+        )
+    """
+    code: str = Field(
+        ..., 
+        min_length=1, 
+        max_length=50,
+        description="标的代码（如 META, 513100.SH, BTC）"
+    )
+    start: Optional[str] = Field(
+        None,
+        pattern=r'^\d{4}-\d{2}-\d{2}$',
+        description="开始日期（YYYY-MM-DD），默认90天前"
+    )
+    end: Optional[str] = Field(
+        None,
+        pattern=r'^\d{4}-\d{2}-\d{2}$',
+        description="结束日期（YYYY-MM-DD），默认今天"
+    )
+    asset_type: Optional[AssetTypeEnum] = Field(
+        None,
+        description="资产类型（可选，覆盖自动检测）"
+    )
+    adj: AdjTypeEnum = Field(
+        AdjTypeEnum.RAW,
+        description="复权类型（raw/qfq/hfq）"
+    )
+    timeframe: Optional[TimeframeEnum] = Field(
+        TimeframeEnum.D1,
+        description="K线周期（仅加密货币需要）"
+    )
+    nocache: bool = Field(
+        False,
+        description="是否跳过缓存"
+    )
+
+    @field_validator('start', 'end')
+    @classmethod
+    def validate_date_format(cls, v: Optional[str]) -> Optional[str]:
+        """验证日期格式"""
+        if v is None:
+            return v
+        
+        # 验证日期合法性
+        from datetime import datetime
+        try:
+            datetime.strptime(v, '%Y-%m-%d')
+        except ValueError:
+            raise ValueError(f'Invalid date format: {v}. Use YYYY-MM-DD')
+        
+        return v
+
+
+class AssetTypeRequest(BaseModel):
+    """资产类型检测请求"""
+    code: str = Field(
+        ...,
+        min_length=1,
+        max_length=50,
+        description="标的代码"
+    )
+
+
+# ============================================================
+# 响应模型
+# ============================================================
+
+class DataRange(BaseModel):
+    """日期范围"""
+    start: Optional[str] = None
+    end: Optional[str] = None
+
+
+class RequestedRange(BaseModel):
+    """用户请求的日期范围"""
+    start: str
+    end: str
+
+
+class AvailableRange(BaseModel):
+    """可用数据范围"""
+    start: Optional[str] = None
+    end: Optional[str] = None
+
+
+class TypeOverride(BaseModel):
+    """类型覆盖信息"""
+    detected: str
+    specified: str
+    hint: str
+
+
+class OHLCVRecord(BaseModel):
+    """
+    OHLCV 数据记录
+    
+    字段会根据资产类型动态变化：
+    - 股票/指数/ETF: date, open, high, low, close, volume
+    - 加密货币: date, open, high, low, close, volume
+    - 其他可能包含: amount, change, pct_chg 等
+    """
+    date: str
+    # 其他字段使用 Dict 接受，因为不同资产类型字段不同
+    
+    class Config:
+        extra = "allow"  # 允许额外字段
+
+
+class NavData(BaseModel):
+    """ETF 净值数据"""
+    data: List[Dict[str, Any]]
+    count: int
+
+
+class PremiumStats(BaseModel):
+    """溢价率统计"""
+    mean: float
+    std: float
+    min: float
+    max: float
+    median: float
+
+
+class PremiumSeriesItem(BaseModel):
+    """溢价率序列项"""
+    date: str
+    premium: float
+
+
+class OHLCVResponse(BaseModel):
+    """
+    OHLCV 数据获取响应
+    
+    字段说明:
+    - code: 标的代码
+    - asset_type: 资产类型
+    - adj: 复权类型
+    - count: 数据条数
+    - data: OHLCV 数据列表
+    - cached: 是否命中缓存
+    - nav: ETF 净值数据（仅 china_etf）
+    - latest_premium: 最新溢价率（仅 china_etf）
+    - premium_series: 溢价率序列（仅 china_etf）
+    - premium_date: 最新溢价率日期（仅 china_etf）
+    - premium_stats: 溢价率统计（仅 china_etf）
+    - attrs: 完整元数据
+    - info: 标的信息（股票/指数）
+    - columns: 数据列名
+    - date_range: 数据日期范围
+    - requested_range: 用户请求范围
+    - available_range: 可用数据范围
+    - cache_strategy: 缓存策略说明
+    - timeframe: K线周期（加密货币）
+    - type_override: 类型覆盖信息
+    
+    使用示例:
+        response = requests.get(url).json()
+        validated = OHLCVResponse.model_validate(response)
+        
+        # 有类型提示
+        print(validated.code)
+        print(validated.latest_premium)
+        print(validated.nav.count if validated.nav else 0)
+    """
+    code: str = Field(..., description="标的代码")
+    asset_type: str = Field(..., description="资产类型")
+    adj: str = Field(..., description="复权类型")
+    count: int = Field(..., description="数据条数")
+    data: List[Dict[str, Any]] = Field(..., description="OHLCV 数据列表")
+    cached: Optional[bool] = Field(None, description="是否命中缓存")
+    
+    # ETF 相关（仅 china_etf）
+    nav: Optional[NavData] = Field(None, description="ETF 净值数据")
+    latest_premium: Optional[float] = Field(None, description="最新溢价率（%）")
+    premium_series: Optional[List[PremiumSeriesItem]] = Field(None, description="溢价率序列")
+    premium_date: Optional[str] = Field(None, description="最新溢价率日期")
+    premium_stats: Optional[PremiumStats] = Field(None, description="溢价率统计")
+    
+    # 元数据
+    attrs: Optional[Dict[str, Any]] = Field(None, description="完整元数据")
+    info: Optional[Dict[str, Any]] = Field(None, description="标的信息")
+    
+    # 辅助信息
+    columns: Optional[List[str]] = Field(None, description="数据列名")
+    date_range: Optional[DataRange] = Field(None, description="数据日期范围")
+    requested_range: Optional[RequestedRange] = Field(None, description="用户请求范围")
+    available_range: Optional[AvailableRange] = Field(None, description="可用数据范围")
+    cache_strategy: Optional[str] = Field(None, description="缓存策略说明")
+    timeframe: Optional[str] = Field(None, description="K线周期（加密货币）")
+    type_override: Optional[TypeOverride] = Field(None, description="类型覆盖信息")
+
+    class Config:
+        extra = "allow"  # 允许额外字段，保持向后兼容
+
+
+class AssetTypeResponse(BaseModel):
+    """资产类型检测响应"""
+    code: str = Field(..., description="标的代码")
+    asset_type: str = Field(..., description="资产类型")
+    description: str = Field(..., description="资产类型描述")
+
+
+class CacheStats(BaseModel):
+    """缓存统计信息"""
+    lru_cache: Dict[str, Any] = Field(..., description="LRU 缓存统计")
+    ttl_cache_size: int = Field(..., description="TTL 缓存大小")
+    ttl_seconds: int = Field(..., description="TTL 秒数")
+    default_start_date: str = Field(..., description="默认数据起点")
+    cache_strategy: str = Field(..., description="缓存策略")
+
+
+class ErrorResponse(BaseModel):
+    """错误响应"""
+    error: str = Field(..., description="错误信息")
+    code: Optional[str] = Field(None, description="标的代码")
+    asset_type: Optional[str] = Field(None, description="资产类型")
+    adj: Optional[str] = Field(None, description="复权类型")
+    detected_type: Optional[str] = Field(None, description="自动检测的类型")
+    example: Optional[str] = Field(None, description="示例请求")
+    hint: Optional[str] = Field(None, description="提示信息")
+    valid_adj: Optional[List[str]] = Field(None, description="有效的 adj 值")
+    valid_types: Optional[List[str]] = Field(None, description="有效的资产类型")
+    valid_timeframes: Optional[List[str]] = Field(None, description="有效的 timeframe 值")
+
+    class Config:
+        extra = "allow"  # 允许额外字段
+
+
+# ============================================================
+# 模型工具函数
+# ============================================================
+
+def validate_ohlcv_response(data: Dict[str, Any]) -> OHLCVResponse:
+    """
+    验证 OHLCV API 响应
+    
+    Args:
+        data: API 返回的 JSON 数据
+    
+    Returns:
+        验证后的响应对象
+    
+    Raises:
+        ValidationError: 数据不符合模型定义
+    
+    使用示例:
+        response = requests.get(url).json()
+        validated = validate_ohlcv_response(response)
+        
+        # 安全访问字段
+        if validated.nav:
+            print(f"净值数据条数: {validated.nav.count}")
+    """
+    return OHLCVResponse.model_validate(data)
+
+
+def dataframe_to_records(df) -> List[Dict[str, Any]]:
+    """
+    将 DataFrame 转换为 OHLCVRecord 兼容的字典列表
+    
+    Args:
+        df: pandas DataFrame
+    
+    Returns:
+        字典列表（可直接用于 OHLCVResponse.data）
+    """
+    if df is None or len(df) == 0:
+        return []
+    
+    df_reset = df.reset_index()
+    
+    # 处理日期列
+    date_columns = ['date', 'Date', 'index', 'trade_date', 'datetime']
+    for col in date_columns:
+        if col in df_reset.columns:
+            try:
+                import pandas as pd
+                df_reset[col] = pd.to_datetime(df_reset[col]).dt.strftime('%Y-%m-%d')
+                if col != 'date':
+                    df_reset = df_reset.rename(columns={col: 'date'})
+                break
+            except Exception:
+                pass
+    
+    return df_reset.to_dict(orient='records')