diff --git a/specs/llm-types-design.md b/specs/llm-types-design.md
new file mode 100644
index 0000000..2a36ac5
--- /dev/null
+++ b/specs/llm-types-design.md
@@ -0,0 +1,243 @@
+# 方案：重构 `types.rs` 为完整的 OpenAI 兼容 API 类型系统
+
+## 1. 现状分析
+
+### 当前问题
+
+| 问题 | 详细 |
+|------|------|
+| **无 serde** | 所有类型只有 `Debug + Clone`，无 `Serialize/Deserialize`，迫使 `OpenaiProvider` 手动构建 JSON（354 行中约 200 行是序列化代码） |
+| **请求参数不全** | `ChatRequest` 只支持 `model, messages, system_prompt, tools, max_tokens, temperature, extra_body`，缺失 streaming、response_format、tool_choice、stop、reasoning_effort 等 30+ 参数 |
+| **响应类型太薄** | `ChatResponse` 只返回 `message + usage + stop_reason`，缺失 `id, created, model, choices` 数组、`logprobs`、`system_fingerprint` 等 |
+| **无流式支持** | 无 `ChatCompletionChunk` 类型，无法处理 SSE 流式响应 |
+| **反向依赖** | `types.rs` 引用 `cycle::usage::Usage`，造成模块间反向依赖 |
+| **手动解析易出错** | `parse_response()` 从 `Value` 中逐字段解析，逻辑脆弱，不支持复杂嵌套类型 |
+
+### OpenAI API 参考文档覆盖范围
+
+已完整阅读文档（2177 行），涵盖了完整的请求参数（35+ 个顶层参数）和响应结构。
+
+## 2. 新类型系统设计
+
+### 架构
+
+将 `types.rs` 重构为 Rust 新风格模块目录（符合项目已有惯例），按功能领域拆分：
+
+```
+src/llm/
+├── types/
+│   ├── mod.rs        # 模块根：re-exports + 基础枚举/共用类型
+│   ├── request.rs    # 请求参数（ChatCompletionRequest 等）
+│   ├── response.rs   # 响应类型（ChatCompletionResponse + ChatCompletionChunk）
+│   ├── message.rs    # 消息类型（6 种角色消息 + content parts）
+│   ├── tool.rs       # 工具定义 + 工具调用
+│   ├── usage.rs      # Token 用量（从 cycle/usage.rs 移入，消除反向依赖）
+│   └── shared.rs     # 共用枚举（ReasoningEffort, ServiceTier, ResponseFormat 等）
+```
+
+同时，将 `cycle/usage.rs` 中的 `Usage` 和 `CostTracker` **移到** `types/usage.rs`，`cycle/usage.rs` 保留 `pub use` 兼容 re-export。
+
+### 核心决策
+
+| 决策 | 选择 | 理由 |
+|------|------|------|
+| **序列化方式** | 全部类型 derive `Serialize, Deserialize` | 消除手动 JSON 构建，让 provider 直接 `.json(&req)` / `.json::<Res>()` |
+| **类型风格** | 直接映射 OpenAI API JSON 形状 | 一目了然，与 API 文档 1:1 对应，调试方便 |
+| **命名策略** | 添加 `OpenAI` 前缀（如 `OpenaiChatRequest`） | 明确标注为 OpenAI 兼容类型 |
+| **字段命名** | `#[serde(rename_all = "snake_case")]` | OpenAI API 使用 snake_case |
+| **可选字段** | `#[serde(skip_serializing_if = "Option::is_none")]` | 不序列化 None 字段，保持请求体干净 |
+| **默认值** | `#[serde(default)]` | 反序列化时缺失字段用默认值 |
+| **后向兼容** | 通过类型别名保持 `ChatRequest`/`ChatResponse` 等名称可用 | LlmProvider/LlmCycle 接口不变 |
+| **泛化策略** | Anthropic 是独立体系，暂不纳入当前设计 | 保持当前类型系统专注 OpenAI，Provider 层做转换 |
+
+### 关键类型设计原则
+
+- **`OpenaiChatRequest`**：统一结构体（不拆分 NonStreaming/Streaming），包含 `stream: Option<bool>` 字段，所有字段均为 `Option`，build 时 `skip_serializing_if`
+- **`OpenaiChatResponse`**：直接对应 `ChatCompletion`（完整响应），保留完整 choices 数组等所有字段
+- **`OpenaiChatChunk`**：对应流式 chunk，`object = "chat.completion.chunk"`
+- **消息系统**：用单个 `OpenaiChatMessage` enum 覆盖 6 种角色消息类型（Developer/System/User/Assistant/Tool/Function），每种内部使用对应 struct
+- **Content parts**：`OpenaiContentPart` enum 覆盖 text/image_url/input_audio/file/refusal
+
+## 3. 完整类型清单
+
+### `types/mod.rs` — 共用类型
+```
+Role                       → enum { Developer, System, User, Assistant, Tool, Function }
+FinishReason               → enum { Stop, Length, ToolCalls, ContentFilter, FunctionCall }
+ServiceTier                → enum { Auto, Default, Flex, Scale, Priority }
+Modality                   → enum { Text, Audio }
+ImageDetail                → enum { Auto, Low, High }
+AudioFormat                → enum { Wav, Mp3, Aac, Flac, Opus, Pcm16 }
+Voice                      → struct { id: String } 或预定义枚举
+SearchContextSize          → enum { Low, Medium, High }
+StopSequence               → enum { Single(String), Multiple(Vec<String>) }
+Verbosity                  → enum { Low, Medium, High }
+```
+
+### `types/request.rs` — 请求参数
+```
+OpenaiChatRequest           → struct (35+ 字段，所有 OpenAI 参数)
+ResponseFormat             → enum { Text, JsonObject { .. }, JsonSchema { .. } }
+ToolChoice                  → enum { None, Auto, Required, Named { .. }, AllowedTools { .. } }
+StreamOptions               → struct { include_usage, include_obfuscation }
+AudioParam                  → struct { format, voice }
+PredictionContent           → struct { type, content }
+WebSearchOptions            → struct { search_context_size, user_location }
+UserLocation                → struct { type, approximate: Approximate }
+Approximate                 → struct { city, country, region, timezone }
+FunctionCallOption          → struct { name }  // deprecated
+FunctionDefinition          → struct { name, description, parameters, strict }
+OpenaiTool                  → enum { Function { .. }, Custom { .. } }
+```
+
+### `types/response.rs` — 响应类型
+```
+OpenaiChatResponse          → struct { id, object, created, model, choices, usage, system_fingerprint, service_tier }
+Choice                       → struct { index, message, finish_reason, logprobs }
+OpenaiChatMessage           → struct { content, refusal, role, tool_calls, function_call, audio, annotations }
+OpenaiChatChunk             → struct { id, object, created, model, choices, usage, system_fingerprint, service_tier }
+ChunkChoice                 → struct { index, delta, logprobs, finish_reason }
+Delta                       → struct { role, content, tool_calls, function_call }
+Logprobs                    → struct { content, refusal }
+TokenLogprob                → struct { token, bytes, logprob, top_logprobs }
+TopLogprob                  → struct { token, bytes, logprob }
+Annotation                  → struct { type, url_citation }
+URLCitation                 → struct { end_index, start_index, title, url }
+OpenaiAudio                 → struct { id, data, expires_at, transcript }
+FunctionCall                → struct { name, arguments }
+OpenaiToolCall              → enum { Function { id, function, type }, Custom { id, custom, type } }
+```
+
+### `types/message.rs` — 消息类型
+```
+OpenaiChatMessage           → enum (覆盖 6 种角色消息)
+  DeveloperMessage          → struct { content, role, name }
+  SystemMessage             → struct { content, role, name }
+  UserMessage               → struct { content, role, name }
+  AssistantMessage          → struct { content, refusal, role, name, tool_calls, function_call, audio }
+  ToolMessage               → struct { content, role, tool_call_id }
+  FunctionMessage           → struct { content, role, name }
+OpenaiContentPart           → enum
+  OpenaiContentPartText     → struct { type, text }
+  OpenaiContentPartImage    → struct { type, image_url: ImageURL }
+  OpenaiContentPartInputAudio → struct { type, input_audio: InputAudio }
+  OpenaiContentPartFile     → struct { type, file: FileData }
+  OpenaiContentPartRefusal  → struct { type, refusal }
+ImageURL                    → struct { url, detail }
+InputAudio                  → struct { data, format }
+FileData                    → struct { file_data, file_id, filename }
+```
+
+### `types/tool.rs` — 工具类型
+```
+OpenaiToolDefinition        → struct { name, description, parameters, strict }
+  （保留 ToolDefinition 别名保持后向兼容，重定义为包含所有字段）
+OpenaiToolCall (在请求中使用) → 见 response.rs 中的定义
+```
+
+### `types/usage.rs` — Token 用量
+```
+Usage                      → struct { prompt_tokens, completion_tokens, total_tokens,
+                                       completion_tokens_details, prompt_tokens_details }
+CompletionTokensDetails    → struct { reasoning_tokens, audio_tokens,
+                                       accepted_prediction_tokens, rejected_prediction_tokens }
+PromptTokensDetails        → struct { audio_tokens, cached_tokens }
+CostTracker                → 从 cycle/usage.rs 移入（累计追踪器）
+```
+
+### 删除的旧类型
+- `ContentBlock` → 被 `OpenaiContentPart` 替代（更准确的 OpenAI API 命名）
+- `StopReason` → 被 `FinishReason` 替代（与 API 命名一致）
+- `Message` → 被 `OpenaiChatMessage` 替代
+
+### 类型别名（后向兼容）
+```
+ChatRequest     = OpenaiChatRequest
+ChatResponse    = OpenaiChatResponse
+Message         = OpenaiChatMessage
+ContentBlock    = OpenaiContentPart
+ToolDefinition  = OpenaiToolDefinition
+Role            = Role（保持不变，但扩展变体）
+StopReason      = FinishReason
+```
+
+## 4. 对其他模块的影响
+
+### `provider/openai.rs`
+- **大幅简化**：`build_request_body()` → 直接 `serde_json::to_value(&request)`
+- `parse_response()` 中 100+ 行手动解析 → 直接 `serde_json::from_value::<OpenaiChatResponse>()`
+- `serialize_messages()`, `serialize_message()`, `serialize_content_block()`, `serialize_tool()` → **全部删除**
+- 新增 `chat_stream()` 方法返回 `OpenaiChatChunk` 流
+- 需要适配新类型的字段名变更（如 `Usage` 中 `input_tokens` → `prompt_tokens`）
+
+### `provider.rs` (trait)
+- 接口保持不变，继续使用 `ChatRequest`/`ChatResponse` 类型别名
+- 调整 `Usage` 类型引用路径
+
+### `cycle.rs`
+- `CycleConfig` 扩展支持更多请求参数（至少增加 `tools, tool_choice, response_format, stop, reasoning_effort, seed` 等）
+- `LlmCycle::submit()` 构建 `ChatRequest` 时使用新类型
+- `response.usage` 字段类型变更（新 `Usage` 含更多字段）
+- 此时不添加流式支持
+
+### `cycle/usage.rs`
+- `Usage` 结构体**被移走**到 `types/usage.rs`
+- `cycle/usage.rs` 保留 `pub use crate::llm::types::usage::{Usage, CostTracker};` 作为兼容性 re-export
+- `CostTracker` 逻辑不变
+
+### `error.rs`
+- 无明显变更，错误类型和映射逻辑不变
+
+## 5. 实施步骤
+
+### Phase 1: 基础设施
+```
+1. [准备] 在 Cargo.toml 中确认 serde 依赖（已有 serde = "1"，features = ["derive"]）
+2. [创建] 新建 src/llm/types/ 目录
+```
+
+### Phase 2: 类型定义（按依赖顺序）
+```
+3. [usage.rs] 从 cycle/usage.rs 迁移 Usage + CostTracker
+4. [shared.rs] 定义 Role, FinishReason, ServiceTier, Modality, ImageDetail, StopSequence, ResponseFormat
+5. [message.rs] 定义 OpenaiChatMessage（6种角色）+ OpenaiContentPart + ImageURL + InputAudio
+6. [tool.rs] 定义 OpenaiToolDefinition + OpenaiToolCall + FunctionCall
+7. [request.rs] 定义 OpenaiChatRequest（35+ 字段）+ ToolChoice + StreamOptions
+8. [response.rs] 定义 OpenaiChatResponse + OpenaiChatChunk + Choice + Delta + Logprobs
+```
+
+### Phase 3: 模块组装
+```
+9. [mod.rs] 创建模块根，re-export 所有类型 + 别名（ChatRequest = OpenaiChatRequest 等）
+10. [usage.rs] 更新 cycle/usage.rs 为 pub use re-export
+11. [删除] 删除旧 src/llm/types.rs
+```
+
+### Phase 4: Provider 适配
+```
+12. [provider/openai.rs] 重写为 serde 序列化（删除 ~200 行手动代码）
+13. [cycle.rs] 适配新类型字段（prompt_tokens vs input_tokens）
+```
+
+### Phase 5: 验证
+```
+14. [编译] cargo check 确保编译通过
+15. [检查] cargo clippy 确保无警告
+16. [测试] cargo test 确保测试通过
+```
+
+## 6. 验证方式
+
+- `cargo check` — 编译通过
+- `cargo clippy` — 无警告
+- `cargo test` — 所有测试通过（如果有集成测试，可能需要调整）
+- 检查 `OpenaiProvider` 代码量减少（预期从 354 行降至 ~150 行）
+- 手动验证序列化输出是否符合 OpenAI API 格式
+
+## 7. 注意事项
+
+1. **Break change**: 某些类型名称变化（如 `StopReason` → `FinishReason`），项目处于早期阶段，可接受
+2. **后向兼容**: 通过类型别名保持旧名称可用，接口层无需修改
+3. **Anthropic 处理**: Anthropic 是独立体系，不在当前设计中泛化，单独实现 Provider
+4. **异步流**: `chat_stream()` 的签名需要仔细设计（`Pin<Box<dyn Stream<Item = Result<OpenaiChatChunk, LlmError>>>>` 或自定义类型）
+5. **CostTracker 不变**: 虽然 Usage 变复杂了，但 CostTracker 只累计 input/output token 数，逻辑不变
diff --git a/src/llm/cycle.rs b/src/llm/cycle.rs
index f710818..250fc73 100644
--- a/src/llm/cycle.rs
+++ b/src/llm/cycle.rs
@@ -7,19 +7,15 @@ pub use usage::{CostTracker, Usage};
 use crate::llm::cycle::retry::should_retry;
 use crate::llm::error::LlmError;
 use crate::llm::provider::LlmProvider;
-use crate::llm::types::{ChatRequest, ChatResponse, ContentBlock, Message, Role, ToolDefinition};
+use crate::llm::types::{
+    ChatRequest, ChatResponse, OpenaiChatMessage, OpenaiTool, ToolChoice, ToolDefinition,
+};
 
-/// LLM 生命周期引擎的配置。
 pub struct CycleConfig {
-    /// 使用的模型名称。
     pub model: String,
-    /// 最大输出 token 数。
     pub max_tokens: Option<u32>,
-    /// 采样温度。
     pub temperature: Option<f32>,
-    /// 最大对话轮数（预留，暂未使用）。
     pub max_turns: Option<u32>,
-    /// 重试策略配置。
     pub retry: RetryConfig,
 }
 
@@ -35,22 +31,15 @@ impl Default for CycleConfig {
     }
 }
 
-/// LLM 调用生命周期引擎。
-///
-/// 管理一次多轮交互的完整生命周期，包括：
-/// - 消息历史维护
-/// - Token 用量追踪
-/// - 自动重试
 pub struct LlmCycle {
     provider: Box<dyn LlmProvider>,
     config: CycleConfig,
     usage: CostTracker,
-    messages: Vec<Message>,
+    messages: Vec<OpenaiChatMessage>,
     system_prompt: Option<String>,
 }
 
 impl LlmCycle {
-    /// 创建新的生命周期引擎。
     pub fn new(provider: Box<dyn LlmProvider>, config: CycleConfig) -> Self {
         Self {
             provider,
@@ -61,50 +50,33 @@ impl LlmCycle {
         }
     }
 
-    /// 设置系统提示词（Builder 模式）。
     pub fn with_system_prompt(mut self, prompt: String) -> Self {
         self.system_prompt = Some(prompt);
         self
     }
 
-    /// 获取 Token 用量追踪器引用。
     pub fn usage(&self) -> &CostTracker {
         &self.usage
     }
 
-    /// 获取当前消息历史。
-    pub fn messages(&self) -> &[Message] {
+    pub fn messages(&self) -> &[OpenaiChatMessage] {
         &self.messages
     }
 
-    /// 清空消息历史。
     pub fn clear_messages(&mut self) {
         self.messages.clear();
     }
 
-    /// 重置 Token 用量统计。
     pub fn reset_usage(&mut self) {
         self.usage.reset();
     }
 
-    /// 提交一条用户消息并获取模型响应。
-    ///
-    /// 流程：
-    /// 1. 将用户消息追加到消息历史
-    /// 2. 构建 ChatRequest
-    /// 3. 使用重试循环调用 provider.chat()
-    /// 4. 将助手回复追加到消息历史
-    /// 5. 累计 token 用量
-    /// 6. 返回 ChatResponse
     pub async fn submit(
         &mut self,
         prompt: String,
         tools: Vec<ToolDefinition>,
     ) -> Result<ChatResponse, LlmError> {
-        self.messages.push(Message {
-            role: Role::User,
-            content: vec![ContentBlock::Text { text: prompt }],
-        });
+        self.messages.push(OpenaiChatMessage::user_text(prompt));
 
         let mut attempts = 0;
 
@@ -113,10 +85,7 @@ impl LlmCycle {
 
             match self.provider.chat(request).await {
                 Ok(response) => {
-                    self.messages.push(Message {
-                        role: Role::Assistant,
-                        content: response.message.content.clone(),
-                    });
+                    self.messages.push(response.message.clone());
 
                     self.usage.add(&response.usage);
 
@@ -134,16 +103,35 @@ impl LlmCycle {
         }
     }
 
-    /// 根据当前状态构建 ChatRequest。
     fn build_request(&self, tools: &[ToolDefinition]) -> ChatRequest {
+        let mut messages = self.messages.clone();
+
+        if let Some(sys_prompt) = &self.system_prompt
+            && !messages.iter().any(|m| matches!(m, OpenaiChatMessage::System { .. }))
+        {
+            messages.insert(0, OpenaiChatMessage::system_text(sys_prompt));
+        }
+
+        let openai_tools: Option<Vec<OpenaiTool>> = if tools.is_empty() {
+            None
+        } else {
+            Some(
+                tools.iter()
+                    .map(|t| OpenaiTool::Function {
+                        function: t.clone(),
+                    })
+                    .collect(),
+            )
+        };
+
         ChatRequest {
             model: self.config.model.clone(),
-            messages: self.messages.clone(),
-            system_prompt: self.system_prompt.clone(),
-            tools: tools.to_vec(),
+            messages,
             max_tokens: self.config.max_tokens,
             temperature: self.config.temperature,
-            extra_body: None,
+            tools: openai_tools,
+            tool_choice: Some(ToolChoice::Auto),
+            ..Default::default()
         }
     }
-}
+}
\ No newline at end of file
diff --git a/src/llm/cycle/usage.rs b/src/llm/cycle/usage.rs
index 32da25e..52974c8 100644
--- a/src/llm/cycle/usage.rs
+++ b/src/llm/cycle/usage.rs
@@ -1,42 +1 @@
-/// 单次请求的 Token 用量。
-#[derive(Debug, Clone, Default)]
-pub struct Usage {
-    /// 输入（提示词）消耗的 token 数。
-    pub input_tokens: u32,
-    /// 输出（生成内容）消耗的 token 数。
-    pub output_tokens: u32,
-}
-
-/// Token 用量累计追踪器。
-///
-/// 在多轮对话中累计所有请求的 token 消耗。
-#[derive(Debug, Default)]
-pub struct CostTracker {
-    accumulated: Usage,
-}
-
-impl CostTracker {
-    /// 累加一次请求的用量。
-    ///
-    /// 使用 saturating_add 防止溢出。
-    pub fn add(&mut self, usage: &Usage) {
-        self.accumulated.input_tokens = self
-            .accumulated
-            .input_tokens
-            .saturating_add(usage.input_tokens);
-        self.accumulated.output_tokens = self
-            .accumulated
-            .output_tokens
-            .saturating_add(usage.output_tokens);
-    }
-
-    /// 获取累计用量。
-    pub fn total(&self) -> &Usage {
-        &self.accumulated
-    }
-
-    /// 重置累计用量。
-    pub fn reset(&mut self) {
-        self.accumulated = Usage::default();
-    }
-}
+pub use crate::llm::types::usage::{CompletionTokensDetails, CostTracker, PromptTokensDetails, Usage, Usage as LlmUsage};
diff --git a/src/llm/provider/openai.rs b/src/llm/provider/openai.rs
index e0b72b1..fb9a605 100644
--- a/src/llm/provider/openai.rs
+++ b/src/llm/provider/openai.rs
@@ -2,33 +2,19 @@ use std::time::Duration;
 
 use async_trait::async_trait;
 use reqwest::Client;
-use serde_json::{json, Value};
 
-use crate::llm::cycle::usage::Usage;
 use crate::llm::error::LlmError;
-use crate::llm::types::{
-    ChatRequest, ChatResponse, ContentBlock, Message, Role, StopReason, ToolDefinition,
-};
-
+use crate::llm::types::{ChatRequest, ChatResponse, OpenaiChatResponse};
 use super::LlmProvider;
 
-/// OpenAI 兼容 API 的 Provider 实现。
-///
-/// 支持任意实现了 `POST /v1/chat/completions` 标准的 API
-/// （包括 OpenAI、Azure OpenAI、DashScope、vLLM 等）。
 pub struct OpenaiProvider {
     http_client: Client,
     base_url: String,
     api_key: String,
-    #[allow(dead_code)]
-    model: String,
 }
 
 impl OpenaiProvider {
-    /// 创建新的 OpenAI Provider。
-    ///
-    /// 默认使用 120 秒超时的 HTTP 客户端。
-    pub fn new(base_url: String, api_key: String, model: String) -> Self {
+    pub fn new(base_url: String, api_key: String, _model: String) -> Self {
         let http_client = Client::builder()
             .timeout(Duration::from_secs(120))
             .build()
@@ -38,284 +24,45 @@ impl OpenaiProvider {
             http_client,
             base_url,
             api_key,
-            model,
         }
     }
 
-    /// 替换为自定义的 HTTP 客户端（用于测试或自定义配置）。
     pub fn with_client(mut self, client: Client) -> Self {
         self.http_client = client;
         self
     }
 
-    /// 将 ChatRequest 构建为 OpenAI API 请求体 JSON。
-    fn build_request_body(&self, request: &ChatRequest) -> Value {
-        let mut body = json!({
-            "model": request.model,
-            "messages": Self::serialize_messages(request),
-        });
-
-        if let Some(max_tokens) = request.max_tokens {
-            body["max_tokens"] = json!(max_tokens);
-        }
-        if let Some(temperature) = request.temperature {
-            body["temperature"] = json!(temperature);
-        }
-        if !request.tools.is_empty() {
-            body["tools"] = json!(
-                request
-                    .tools
-                    .iter()
-                    .map(Self::serialize_tool)
-                    .collect::<Vec<_>>()
-            );
-        }
-
-        // 合并 extra_body 中的扩展参数到请求体顶层
-        if let Some(ref extra) = request.extra_body
-            && let Some(obj) = extra.as_object()
-        {
-            for (k, v) in obj {
-                body[k] = v.clone();
+    fn map_reqwest_error(e: reqwest::Error) -> LlmError {
+        if e.is_timeout() {
+            LlmError::Timeout {
+                duration: Duration::from_secs(120),
             }
-        }
-
-        body
-    }
-
-    /// 将请求中的消息列表序列化为 API 消息数组。
-    fn serialize_messages(request: &ChatRequest) -> Vec<Value> {
-        let mut messages: Vec<Value> = Vec::new();
-
-        // system_prompt 作为独立的 system 角色消息放在最前面
-        if let Some(ref system_prompt) = request.system_prompt {
-            messages.push(json!({
-                "role": "system",
-                "content": system_prompt
-            }));
-        }
-
-        for msg in &request.messages {
-            messages.push(Self::serialize_message(msg));
-        }
-
-        messages
-    }
-
-    /// 将单条消息序列化为 API 格式。
-    ///
-    /// 处理逻辑：
-    /// - 多个 content block 或包含图片 → 使用数组格式
-    /// - ToolResult → 使用 tool 角色格式
-    /// - 其他 → 使用纯文本格式
-    fn serialize_message(msg: &Message) -> Value {
-        let role_str = match msg.role {
-            Role::User => "user",
-            Role::Assistant => "assistant",
-            Role::System => "system",
-            Role::Tool => "tool",
-        };
-
-        let has_mixed = msg.content.len() > 1
-            || msg
-                .content
-                .iter()
-                .any(|b| matches!(b, ContentBlock::ImageUrl { .. }));
-
-        if has_mixed {
-            let content: Vec<Value> = msg
-                .content
-                .iter()
-                .map(Self::serialize_content_block)
-                .collect();
-            json!({ "role": role_str, "content": content })
-        } else if let Some(ContentBlock::ToolResult {
-            tool_use_id,
-            content,
-        }) = msg.content.first()
-        {
-            json!({
-                "role": "tool",
-                "tool_call_id": tool_use_id,
-                "content": content
-            })
+        } else if e.is_connect() {
+            LlmError::Other(format!("连接失败: {}", e))
         } else {
-            let text = msg
-                .content
-                .first()
-                .map(|b| match b {
-                    ContentBlock::Text { text } => text.clone(),
-                    _ => String::new(),
-                })
-                .unwrap_or_default();
-            json!({ "role": role_str, "content": text })
+            LlmError::Other(format!("请求失败: {}", e))
         }
     }
-
-    /// 将 ContentBlock 序列化为 API content parts 数组元素。
-    fn serialize_content_block(block: &ContentBlock) -> Value {
-        match block {
-            ContentBlock::Text { text } => {
-                json!({ "type": "text", "text": text })
-            }
-            ContentBlock::ImageUrl { url } => {
-                json!({ "type": "image_url", "image_url": { "url": url } })
-            }
-            ContentBlock::ToolUse { id, name, input } => {
-                json!({ "type": "tool_use", "id": id, "name": name, "input": input })
-            }
-            ContentBlock::ToolResult { .. } => {
-                json!({ "type": "tool_result", "content": "" })
-            }
-        }
-    }
-
-    /// 将 ToolDefinition 序列化为 OpenAI tools 数组元素。
-    fn serialize_tool(tool: &ToolDefinition) -> Value {
-        json!({
-            "type": "function",
-            "function": {
-                "name": tool.name,
-                "description": tool.description,
-                "parameters": tool.input_schema
-            }
-        })
-    }
-
-    /// 将 OpenAI API 响应 JSON 解析为 ChatResponse。
-    fn parse_response(response: Value) -> Result<ChatResponse, LlmError> {
-        let choice = response["choices"][0]
-            .as_object()
-            .ok_or_else(|| LlmError::Other("响应中缺少 choices[0]".into()))?;
-
-        let msg = choice["message"]
-            .as_object()
-            .ok_or_else(|| LlmError::Other("响应中缺少 message".into()))?;
-
-        let role = match msg["role"].as_str() {
-            Some("assistant") => Role::Assistant,
-            Some(_) => Role::Assistant,
-            None => Role::Assistant,
-        };
-
-        let mut content_blocks: Vec<ContentBlock> = Vec::new();
-
-        // 从 content 字段提取文本和 tool_use
-        if let Some(content_val) = msg.get("content") {
-            match content_val {
-                Value::String(s) if !s.is_empty() => {
-                    content_blocks.push(ContentBlock::Text { text: s.clone() });
-                }
-                Value::Array(arr) => {
-                    for item in arr {
-                        if let Some(item_type) = item["type"].as_str() {
-                            match item_type {
-                                "text" => {
-                                    if let Some(text) = item["text"].as_str() {
-                                        content_blocks
-                                            .push(ContentBlock::Text { text: text.into() });
-                                    }
-                                }
-                                "tool_use" | "function" => {
-                                    let id = item["id"].as_str().unwrap_or("").to_string();
-                                    let name = item["name"].as_str().unwrap_or("").to_string();
-                                    let input = item.get("input").cloned().unwrap_or(Value::Null);
-                                    content_blocks
-                                        .push(ContentBlock::ToolUse { id, name, input });
-                                }
-                                _ => {}
-                            }
-                        }
-                    }
-                }
-                _ => {}
-            }
-        }
-
-        // 从 tool_calls 字段提取工具调用（OpenAI 特有格式）
-        if let Some(tool_calls) = msg.get("tool_calls").and_then(|v| v.as_array()) {
-            for tc in tool_calls {
-                let id = tc["id"].as_str().unwrap_or("").to_string();
-                let name = tc["function"]["name"].as_str().unwrap_or("").to_string();
-                let input = tc["function"]["arguments"]
-                    .as_str()
-                    .and_then(|s| serde_json::from_str(s).ok())
-                    .unwrap_or(Value::Null);
-                content_blocks.push(ContentBlock::ToolUse { id, name, input });
-            }
-        }
-
-        if content_blocks.is_empty() {
-            content_blocks.push(ContentBlock::Text {
-                text: String::new(),
-            });
-        }
-
-        // 解析停止原因
-        let stop_reason = choice["finish_reason"].as_str().map(|s| match s {
-            "stop" => StopReason::Stop,
-            "tool_calls" => StopReason::ToolUse,
-            "max_tokens" => StopReason::MaxTokens,
-            "length" => StopReason::Length,
-            "content_filter" => StopReason::ContentFilter,
-            other => StopReason::Other(other.into()),
-        });
-
-        // 解析 token 用量
-        let usage = response["usage"]
-            .as_object()
-            .map(|u| Usage {
-                input_tokens: u.get("prompt_tokens").and_then(|v| v.as_u64()).unwrap_or(0) as u32,
-                output_tokens: u
-                    .get("completion_tokens")
-                    .and_then(|v| v.as_u64())
-                    .unwrap_or(0) as u32,
-            })
-            .unwrap_or_default();
-
-        Ok(ChatResponse {
-            message: Message {
-                role,
-                content: content_blocks,
-            },
-            usage,
-            stop_reason,
-        })
-    }
 }
 
 #[async_trait]
 impl LlmProvider for OpenaiProvider {
     async fn chat(&self, request: ChatRequest) -> Result<ChatResponse, LlmError> {
         let url = format!("{}/chat/completions", self.base_url.trim_end_matches('/'));
-        let body = self.build_request_body(&request);
 
         let response = self
             .http_client
             .post(&url)
             .header("Authorization", format!("Bearer {}", self.api_key))
-            .header("Content-Type", "application/json")
-            .json(&body)
+            .json(&request)
             .send()
             .await
-            .map_err(|e| {
-                if e.is_timeout() {
-                    LlmError::Timeout {
-                        duration: Duration::from_secs(120),
-                    }
-                } else if e.is_connect() {
-                    LlmError::Other(format!("连接失败: {}", e))
-                } else {
-                    LlmError::Other(format!("请求失败: {}", e))
-                }
-            })?;
+            .map_err(Self::map_reqwest_error)?;
 
         let status = response.status();
         let status_code: u16 = status.as_u16();
 
-        // 处理非 2xx 响应，将 HTTP 状态码映射为对应的 LlmError 变体
         if !status.is_success() {
-            // 在消费 response body 之前先读取 retry-after 头部
             let retry_after = response
                 .headers()
                 .get("retry-after")
@@ -344,11 +91,11 @@ impl LlmProvider for OpenaiProvider {
             };
         }
 
-        let json_body: Value = response
+        let chat_response: OpenaiChatResponse = response
             .json()
             .await
             .map_err(|e| LlmError::Other(format!("响应解析失败: {}", e)))?;
 
-        Self::parse_response(json_body)
+        Ok(ChatResponse::from(chat_response))
     }
-}
+}
\ No newline at end of file
diff --git a/src/llm/types.rs b/src/llm/types.rs
deleted file mode 100644
index 950ac44..0000000
--- a/src/llm/types.rs
+++ /dev/null
@@ -1,100 +0,0 @@
-use crate::llm::cycle::usage::Usage;
-use serde_json::Value;
-
-/// 对话消息的角色。
-#[derive(Debug, Clone)]
-pub enum Role {
-    User,
-    Assistant,
-    System,
-    Tool,
-}
-
-/// 消息内容块，支持多模态及工具调用。
-#[derive(Debug, Clone)]
-pub enum ContentBlock {
-    /// 纯文本内容。
-    Text {
-        text: String,
-    },
-    /// 图片 URL（多模态输入预留）。
-    ImageUrl {
-        url: String,
-    },
-    /// 模型发起的工具调用（预留，暂不实现自动执行）。
-    ToolUse {
-        id: String,
-        name: String,
-        input: Value,
-    },
-    /// 工具执行结果的回传（预留，暂不实现自动执行）。
-    ToolResult {
-        tool_use_id: String,
-        content: String,
-    },
-}
-
-/// 一条对话消息，由角色和内容块列表组成。
-#[derive(Debug, Clone)]
-pub struct Message {
-    pub role: Role,
-    pub content: Vec<ContentBlock>,
-}
-
-/// 可供模型调用的工具定义。
-#[derive(Debug, Clone)]
-pub struct ToolDefinition {
-    /// 工具名称。
-    pub name: String,
-    /// 工具描述，用于模型理解何时调用。
-    pub description: String,
-    /// JSON Schema 格式的输入参数定义。
-    pub input_schema: Value,
-}
-
-/// 对 /v1/chat/completions 的完整请求参数。
-#[derive(Debug, Clone)]
-pub struct ChatRequest {
-    /// 模型标识（如 "gpt-4o"）。
-    pub model: String,
-    /// 对话历史 + 新消息。
-    pub messages: Vec<Message>,
-    /// 独立的系统提示词，将在序列化时转为 system 角色消息。
-    pub system_prompt: Option<String>,
-    /// 可用的工具定义列表。
-    pub tools: Vec<ToolDefinition>,
-    /// 最大输出 token 数。
-    pub max_tokens: Option<u32>,
-    /// 采样温度。
-    pub temperature: Option<f32>,
-    /// 扩展参数（如 enable_thinking），会合并到请求体顶层。
-    pub extra_body: Option<Value>,
-}
-
-/// 模型返回的完整响应。
-#[derive(Debug, Clone)]
-pub struct ChatResponse {
-    /// 助手的回复消息。
-    pub message: Message,
-    /// 本次请求的 token 用量。
-    pub usage: Usage,
-    /// 停止原因。
-    pub stop_reason: Option<StopReason>,
-}
-
-/// 模型停止生成的原因。
-#[derive(Debug, Clone)]
-pub enum StopReason {
-    /// 正常结束。
-    Stop,
-    /// 模型请求调用工具（预留）。
-    ToolUse,
-    /// 达到 max_tokens 上限。
-    MaxTokens,
-    /// 内容被安全过滤。
-    ContentFilter,
-    /// 长度限制（兼容某些 API 的 finish_reason）。
-    Length,
-    /// 其他未分类的原因。
-    Other(String),
-}
diff --git a/src/llm/types/message.rs b/src/llm/types/message.rs
new file mode 100644
index 0000000..d81a7f3
--- /dev/null
+++ b/src/llm/types/message.rs
@@ -0,0 +1,125 @@
+use crate::llm::types::shared::{AudioFormat, ImageDetail};
+use crate::llm::types::tool::OpenaiToolCall;
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct ImageURL {
+    pub url: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub detail: Option<ImageDetail>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct InputAudio {
+    pub data: String,
+    pub format: AudioFormat,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct FileData {
+    pub file_data: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub file_id: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub filename: Option<String>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "type")]
+pub enum OpenaiContentPart {
+    Text {
+        text: String,
+    },
+    Image {
+        image_url: ImageURL,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        detail: Option<ImageDetail>,
+    },
+    InputAudio {
+        input_audio: InputAudio,
+    },
+    File {
+        file: FileData,
+    },
+    Refusal {
+        refusal: String,
+    },
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "role")]
+pub enum OpenaiChatMessage {
+    Developer {
+        content: Vec<OpenaiContentPart>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        name: Option<String>,
+    },
+    System {
+        content: Vec<OpenaiContentPart>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        name: Option<String>,
+    },
+    User {
+        content: Vec<OpenaiContentPart>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        name: Option<String>,
+    },
+    Assistant {
+        content: Vec<OpenaiContentPart>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        refusal: Option<String>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        name: Option<String>,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        tool_calls: Option<Vec<OpenaiToolCall>>,
+    },
+    Tool {
+        content: Vec<OpenaiContentPart>,
+        tool_call_id: String,
+    },
+    Function {
+        content: Vec<OpenaiContentPart>,
+        name: String,
+    },
+}
+
+impl OpenaiChatMessage {
+    pub fn user_text<S: Into<String>>(text: S) -> Self {
+        OpenaiChatMessage::User {
+            content: vec![OpenaiContentPart::Text { text: text.into() }],
+            name: None,
+        }
+    }
+
+    pub fn assistant_text<S: Into<String>>(text: S) -> Self {
+        OpenaiChatMessage::Assistant {
+            content: vec![OpenaiContentPart::Text { text: text.into() }],
+            refusal: None,
+            name: None,
+            tool_calls: None,
+        }
+    }
+
+    pub fn system_text<S: Into<String>>(text: S) -> Self {
+        OpenaiChatMessage::System {
+            content: vec![OpenaiContentPart::Text { text: text.into() }],
+            name: None,
+        }
+    }
+
+    pub fn developer_text<S: Into<String>>(text: S) -> Self {
+        OpenaiChatMessage::Developer {
+            content: vec![OpenaiContentPart::Text { text: text.into() }],
+            name: None,
+        }
+    }
+
+    pub fn tool_result<S: Into<String>>(tool_call_id: String, content: S) -> Self {
+        OpenaiChatMessage::Tool {
+            content: vec![OpenaiContentPart::Text {
+                text: content.into(),
+            }],
+            tool_call_id,
+        }
+    }
+}
diff --git a/src/llm/types/mod.rs b/src/llm/types/mod.rs
new file mode 100644
index 0000000..b3747a9
--- /dev/null
+++ b/src/llm/types/mod.rs
@@ -0,0 +1,53 @@
+pub mod message;
+pub mod request;
+pub mod response;
+pub mod shared;
+pub mod tool;
+pub mod usage;
+
+pub use message::{
+    FileData, ImageURL, InputAudio, OpenaiChatMessage, OpenaiContentPart,
+};
+pub use request::{OpenaiChatRequest, OpenaiTool, StreamOptions, ToolChoice};
+pub use response::{
+    Annotation, Choice, ChunkChoice, Delta, Logprobs, OpenaiAudio,
+    OpenaiChatChunk, OpenaiChatResponse, TokenLogprob, TopLogprob, URLCitation,
+};
+pub use shared::{
+    AudioFormat, FinishReason, ImageDetail, Modality, ResponseFormat, Role,
+    ServiceTier, StopSequence,
+};
+pub use tool::{FunctionCall, OpenaiToolCall, OpenaiToolDefinition};
+pub use usage::{CompletionTokensDetails, CostTracker, PromptTokensDetails, Usage};
+
+#[derive(Debug, Clone)]
+pub struct ChatResponse {
+    pub message: OpenaiChatMessage,
+    pub usage: Usage,
+    pub stop_reason: Option<FinishReason>,
+}
+
+impl From<OpenaiChatResponse> for ChatResponse {
+    fn from(response: OpenaiChatResponse) -> Self {
+        let message = response
+            .choices
+            .first()
+            .map(|c| c.message.clone())
+            .unwrap_or_else(|| OpenaiChatMessage::assistant_text(""));
+        let stop_reason = response
+            .choices
+            .first()
+            .and_then(|c| c.finish_reason);
+        ChatResponse {
+            message,
+            usage: response.usage,
+            stop_reason,
+        }
+    }
+}
+
+pub type ChatRequest = OpenaiChatRequest;
+pub type Message = OpenaiChatMessage;
+pub type ContentBlock = OpenaiContentPart;
+pub type ToolDefinition = OpenaiToolDefinition;
+pub type StopReason = FinishReason;
\ No newline at end of file
diff --git a/src/llm/types/request.rs b/src/llm/types/request.rs
new file mode 100644
index 0000000..5530b9f
--- /dev/null
+++ b/src/llm/types/request.rs
@@ -0,0 +1,117 @@
+use serde::{Deserialize, Serialize};
+use serde_json::Value;
+use crate::llm::types::shared::{ResponseFormat, ServiceTier, StopSequence};
+use crate::llm::types::tool::OpenaiToolDefinition;
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct StreamOptions {
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub include_usage: Option<bool>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub include_obfuscation: Option<bool>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "type")]
+pub enum ToolChoice {
+    None,
+    Auto,
+    Required,
+    Named {
+        name: String,
+    },
+    AllowedTools {
+        #[serde(rename = "tools")]
+        tool_names: Vec<String>,
+    },
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "type")]
+pub enum OpenaiTool {
+    Function {
+        function: OpenaiToolDefinition,
+    },
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct AudioParam {
+    pub format: String,
+    pub voice: String,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct PredictionContent {
+    #[serde(rename = "type")]
+    pub pred_type: String,
+    pub content: String,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct UserLocation {
+    #[serde(rename = "type")]
+    pub loc_type: String,
+    pub approximate: Approximate,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Approximate {
+    pub city: String,
+    pub country: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub region: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub timezone: Option<String>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct WebSearchOptions {
+    pub search_context_size: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub user_location: Option<UserLocation>,
+}
+
+#[derive(Debug, Clone, Default, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub struct OpenaiChatRequest {
+    pub model: String,
+    pub messages: Vec<crate::llm::types::message::OpenaiChatMessage>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub frequency_penalty: Option<f32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub logit_bias: Option<Value>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub max_tokens: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub n: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub presence_penalty: Option<f32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub response_format: Option<ResponseFormat>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub seed: Option<i64>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub service_tier: Option<ServiceTier>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub stop: Option<StopSequence>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub stream: Option<bool>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub stream_options: Option<StreamOptions>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub temperature: Option<f32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub top_p: Option<f32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub tools: Option<Vec<OpenaiTool>>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub tool_choice: Option<ToolChoice>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub parallel_tool_calls: Option<bool>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub user: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub extra_headers: Option<Value>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub extra_body: Option<Value>,
+}
\ No newline at end of file
diff --git a/src/llm/types/response.rs b/src/llm/types/response.rs
new file mode 100644
index 0000000..90011ce
--- /dev/null
+++ b/src/llm/types/response.rs
@@ -0,0 +1,117 @@
+use serde::{Deserialize, Serialize};
+use crate::llm::types::shared::{FinishReason, ServiceTier};
+use crate::llm::types::message::OpenaiChatMessage;
+use crate::llm::types::tool::OpenaiToolCall;
+use crate::llm::types::usage::Usage;
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct TokenLogprob {
+    pub token: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub bytes: Option<Vec<u32>>,
+    pub logprob: f64,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub top_logprobs: Option<Vec<TopLogprob>>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct TopLogprob {
+    pub token: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub bytes: Option<Vec<u32>>,
+    pub logprob: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Logprobs {
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub content: Option<Vec<TokenLogprob>>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub refusal: Option<Vec<TokenLogprob>>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct URLCitation {
+    pub end_index: u32,
+    pub start_index: u32,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub title: Option<String>,
+    pub url: String,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Annotation {
+    #[serde(rename = "type")]
+    pub ann_type: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub url_citation: Option<URLCitation>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct OpenaiAudio {
+    pub id: String,
+    pub data: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub expires_at: Option<i64>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub transcript: Option<String>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Choice {
+    pub index: u32,
+    pub message: OpenaiChatMessage,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub finish_reason: Option<FinishReason>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub logprobs: Option<Logprobs>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct OpenaiChatResponse {
+    pub id: String,
+    pub object: String,
+    pub created: u64,
+    pub model: String,
+    pub choices: Vec<Choice>,
+    pub usage: Usage,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub system_fingerprint: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub service_tier: Option<ServiceTier>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Delta {
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub role: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub content: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub refusal: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub tool_calls: Option<Vec<OpenaiToolCall>>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct ChunkChoice {
+    pub index: u32,
+    pub delta: Delta,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub logprobs: Option<Logprobs>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub finish_reason: Option<FinishReason>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct OpenaiChatChunk {
+    pub id: String,
+    pub object: String,
+    pub created: u64,
+    pub model: String,
+    pub choices: Vec<ChunkChoice>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub usage: Option<Usage>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub system_fingerprint: Option<String>,
+}
\ No newline at end of file
diff --git a/src/llm/types/shared.rs b/src/llm/types/shared.rs
new file mode 100644
index 0000000..210e0fc
--- /dev/null
+++ b/src/llm/types/shared.rs
@@ -0,0 +1,78 @@
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum Role {
+    Developer,
+    System,
+    User,
+    Assistant,
+    Tool,
+    Function,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum FinishReason {
+    Stop,
+    Length,
+    ToolCalls,
+    ContentFilter,
+    FunctionCall,
+    #[serde(other)]
+    Other,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum ServiceTier {
+    Auto,
+    Default,
+    #[serde(other)]
+    Other,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum Modality {
+    Text,
+    Audio,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum ImageDetail {
+    Auto,
+    Low,
+    High,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum AudioFormat {
+    Wav,
+    Mp3,
+    Aac,
+    Flac,
+    Opus,
+    Pcm16,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum StopSequence {
+    Single(String),
+    Multiple(Vec<String>),
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "type")]
+pub enum ResponseFormat {
+    Text,
+    JsonObject,
+    JsonSchema {
+        schema: serde_json::Value,
+        #[serde(skip_serializing_if = "Option::is_none")]
+        strict: Option<bool>,
+    },
+}
\ No newline at end of file
diff --git a/src/llm/types/tool.rs b/src/llm/types/tool.rs
new file mode 100644
index 0000000..1506361
--- /dev/null
+++ b/src/llm/types/tool.rs
@@ -0,0 +1,27 @@
+use serde::{Deserialize, Serialize};
+use serde_json::Value;
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct OpenaiToolDefinition {
+    pub name: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub description: Option<String>,
+    pub parameters: Value,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub strict: Option<bool>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct FunctionCall {
+    pub name: String,
+    pub arguments: String,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "snake_case", tag = "type")]
+pub enum OpenaiToolCall {
+    Function {
+        id: String,
+        function: FunctionCall,
+    },
+}
\ No newline at end of file
diff --git a/src/llm/types/usage.rs b/src/llm/types/usage.rs
new file mode 100644
index 0000000..e7121f6
--- /dev/null
+++ b/src/llm/types/usage.rs
@@ -0,0 +1,75 @@
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, Copy, Default, Serialize, Deserialize)]
+pub struct Usage {
+    pub prompt_tokens: u32,
+    pub completion_tokens: u32,
+    pub total_tokens: u32,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub completion_tokens_details: Option<CompletionTokensDetails>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub prompt_tokens_details: Option<PromptTokensDetails>,
+}
+
+#[derive(Debug, Clone, Copy, Default, Serialize, Deserialize)]
+pub struct CompletionTokensDetails {
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub reasoning_tokens: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub audio_tokens: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub accepted_prediction_tokens: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub rejected_prediction_tokens: Option<u32>,
+}
+
+#[derive(Debug, Clone, Copy, Default, Serialize, Deserialize)]
+pub struct PromptTokensDetails {
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub audio_tokens: Option<u32>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub cached_tokens: Option<u32>,
+}
+
+#[derive(Debug, Default)]
+pub struct CostTracker {
+    accumulated: Usage,
+}
+
+impl CostTracker {
+    pub fn add(&mut self, usage: &Usage) {
+        self.accumulated.prompt_tokens = self
+            .accumulated
+            .prompt_tokens
+            .saturating_add(usage.prompt_tokens);
+        self.accumulated.completion_tokens = self
+            .accumulated
+            .completion_tokens
+            .saturating_add(usage.completion_tokens);
+        self.accumulated.total_tokens = self
+            .accumulated
+            .total_tokens
+            .saturating_add(usage.total_tokens);
+    }
+
+    pub fn total(&self) -> &Usage {
+        &self.accumulated
+    }
+
+    pub fn reset(&mut self) {
+        self.accumulated = Usage::default();
+    }
+}
+
+impl Usage {
+    pub fn from_input_output(input: u32, output: u32) -> Self {
+        let total = input.saturating_add(output);
+        Usage {
+            prompt_tokens: input,
+            completion_tokens: output,
+            total_tokens: total,
+            completion_tokens_details: None,
+            prompt_tokens_details: None,
+        }
+    }
+}
\ No newline at end of file