diff --git a/CHANGELOG.md b/CHANGELOG.md
index ca9dd7e..f63ff74 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -18,7 +18,11 @@ This package is used in 100k+ production installs. We take backward compatibilit
 - No published interface ([`ClientContract`](src/Contracts/ClientContract.php), `ResourceContract`, `ResultContract`, `ApiFactoryContract`) will gain new required methods. New methods on the client implementation will be exposed via separate, additive interfaces.
 - Default behavior of existing methods will not silently change in ways that affect cost, output, or correctness (e.g. raising default `max_tokens`).
 
-**What's coming in `v2.1.x`:** Every new DeepSeek API feature (V4 models, thinking mode, FIM completion, Anthropic format, user balance, `stop` / `top_p` / `tool_choice` / `logprobs` / `user_id`, chat prefix completion, etc.) will be delivered as **additive** new methods, optional parameters, and enum cases. Bug fixes (`/v3` base URL, ignored params in `chat()` / `code()`, keep-alive line stripping) are also in scope.
+**What's been delivered in the `v2.x` line so far:**
+- **v2.1.0** — V4 models, base-URL fix, `chat()` / `code()` parameter wiring, keep-alive line stripping.
+- **v2.2.0** — thinking mode (`setThinking`, `setReasoningEffort`), `setStop`, `setTopP`, `setToolChoice` (including the previously missing `"required"` mode), `setLogprobs`, `setTopLogprobs`, `setUserId`, OpenAI-spec `name` field on messages, and the `setStrictTool` helper.
+
+**What's still coming in the `v2.x` line:** FIM completion, real SSE streaming, chat prefix completion (Beta), user balance endpoint, rate-limit handling, response-introspection accessors, and additional DX helpers — all additive.
 
 **What's reserved for `v3.0.0`:** Removing deprecated `Models::CODER` / `Models::R1` / `Models::R1Zero`, removing the `Coder` class and `HasCoder` trait, raising default `max_tokens`, retyping `run()` to return a structured DTO, expanding `ClientContract` to match the implementation. All of these will be announced via `@deprecated` notices throughout `v2.x` and shipped with a complete migration guide in [MIGRATION.md](MIGRATION.md).
 
@@ -26,16 +30,12 @@ See [TODO.md](TODO.md) for the full feature gap analysis with per-item BC classi
 
 ---
 
-## [Unreleased] - v2.2.0 (planned)
+## [Unreleased] - v2.2.x (planned follow-ups)
 
 > Additive only. Zero breaking changes from v2.x. See [TODO.md](TODO.md) for the source list.
 
-### Generation parameters and DX
-- Thinking mode setters: `setThinking(array $config)` and `setReasoningEffort(string $effort)`.
-- New sampling / generation parameter setters: `setStop()`, `setTopP()`, `setToolChoice()`, `setLogprobs()`, `setTopLogprobs()`, `setUserId()`.
-- Optional `?string $name` parameter on `query()` and `buildQuery()` for the OpenAI message `name` field.
+### Still planned for the v2.2.x line
 - `setSystemMessage(string $content)` convenience method.
-- Tool `strict` mode helper for function definitions.
 - Response introspection accessors on `SuccessResult`: `getMessage()`, `getUsage()`, `getReasoningContent()`, `getToolCalls()`, `getFinishReason()`, `getCacheHitTokens()` — `run()` continues to return `string`.
 - `getQueries()`, `getConfig()`, `reset()` introspection helpers.
 - `DefaultConfigs::MAX_TOKENS` and `DefaultConfigs::RESPONSE_FORMAT_TYPE` cases (`TemperatureValues::MAX_TOKENS` / `RESPONSE_FORMAT_TYPE` deprecated).
@@ -54,6 +54,50 @@ See [TODO.md](TODO.md) for the full feature gap analysis with per-item BC classi
 
 ---
 
+## [2.2.0] - 2026-05-25
+
+> Additive parameter setters. Zero breaking changes from `v2.1.x`. Existing callers who do not opt into any of the new setters receive a byte-identical request body compared to v2.1.x — this is enforced by the regression tests in [`tests/Feature/V220ChangesTest.php`](tests/Feature/V220ChangesTest.php).
+
+### Added
+
+#### Generation parameter setters (all optional, all omitted from the request body until explicitly invoked)
+
+- **Thinking mode** (`TODO.md` #4) — [`setThinking(array $config)`](src/Traits/Client/HasGenerationParams.php) and [`setReasoningEffort(string $effort)`](src/Traits/Client/HasGenerationParams.php). See the [DeepSeek reasoning_model docs](https://api-docs.deepseek.com/guides/reasoning_model).
+- **Stop sequences** (`TODO.md` #5) — `setStop(string|array $stop)`. Single strings are normalized to a one-element array.
+- **Nucleus sampling** (`TODO.md` #6) — `setTopP(float $topP)`.
+- **Tool choice** (`TODO.md` #7) — `setToolChoice(string|array $toolChoice)`. Accepts `"none"`, `"auto"`, `"required"`, or the named-function array shape `['type' => 'function', 'function' => ['name' => '...']]`. The previously missing `"required"` mode is now reachable.
+- **Log probabilities** (`TODO.md` #8) — `setLogprobs(bool $enabled)` and `setTopLogprobs(int $count)`.
+- **End-user identifier** (`TODO.md` #9) — `setUserId(string $userId)`. Sent on the wire as the OpenAI-spec `user` field.
+
+#### Message-level additions
+
+- **OpenAI-spec `name` field on messages** (`TODO.md` #10) — optional 3rd parameter `?string $name = null` added to [`DeepSeekClient::query()`](src/DeepSeekClient.php) and [`DeepSeekClient::buildQuery()`](src/DeepSeekClient.php). The `name` key is omitted entirely from the message when null, preserving the existing 2-argument behavior byte-for-byte.
+
+#### Function-calling additions
+
+- **Structured tool `strict` mode helper** (`TODO.md` #13) — new [`setStrictTool(string $name, array $parameters, ?string $description = null)`](src/Traits/Client/HasToolsFunctionCalling.php) appends a function tool with `strict: true` to the existing `tools` array. The existing `setTools(array)` API is unchanged.
+
+#### Enums and helpers
+
+- New `QueryFlags` cases: `STOP`, `TOP_P`, `TOOL_CHOICE`, `LOGPROBS`, `TOP_LOGPROBS`, `USER`, `THINKING`, `REASONING_EFFORT`.
+- New [`DeepSeek\Enums\Configs\ReasoningEffort`](src/Enums/Configs/ReasoningEffort.php) enum with `HIGH` and `MAX` cases.
+- New [`DeepSeek\Enums\Configs\ThinkingType`](src/Enums/Configs/ThinkingType.php) enum with `ENABLED` and `DISABLED` cases.
+- New [`DeepSeek\Enums\Queries\ToolChoiceMode`](src/Enums/Queries/ToolChoiceMode.php) enum with `NONE`, `AUTO`, and `REQUIRED` cases.
+- New [`DeepSeek\Traits\Client\HasGenerationParams`](src/Traits/Client/HasGenerationParams.php) trait composed into [`DeepSeekClient`](src/DeepSeekClient.php). Holds all eight new setters and their backing nullable state.
+
+### Internal
+
+- `run()`, `chat()`, `code()` now merge the new optional parameters via an omit-when-null loop. The original seven request-body keys (`messages`, `model`, `stream`, `temperature`, `max_tokens`, `tools`, `response_format`) remain in their original order and are sent unchanged regardless of whether any new setter is called. This is verified by the three "byte-identical when no new setters called" tests at the top of [`tests/Feature/V220ChangesTest.php`](tests/Feature/V220ChangesTest.php).
+- 23 new tests cover BC guards, every new setter, the `name` field, the strict-tool helper, and the new enum surface. The existing test suite (DeepSeekClientTest, V210ChangesTest, FunctionCallingTest) continues to pass unchanged.
+
+### Backward-compatibility notes
+
+- No public method, class, trait, enum case, or constant was removed or renamed.
+- No existing method signature changed in a breaking way. The new optional `?string $name` parameter on `query()` / `buildQuery()` follows the same additive pattern as the `?string $clientType = null` parameter added to `build()` in v2.1.0.
+- `ClientContract` is untouched; the new setters are introduced via the additive `HasGenerationParams` trait on the concrete client.
+
+---
+
 ## [2.1.0] - 2026-05-22
 
 > Foundation + Bug Fixes. Zero breaking changes from `v2.0.x`. All deprecated symbols remain fully functional throughout the `v2.x` line.
diff --git a/README-AR.md b/README-AR.md
index ffa347a..4b35cf3 100644
--- a/README-AR.md
+++ b/README-AR.md
@@ -32,6 +32,15 @@
   - [الاستخدام مع عميل HTTP من Symfony](#الاستخدام-مع-عميل-http-من-symfony)
   - [الحصول على قائمة النماذج](#الحصول-على-قائمة-النماذج)
   - [استدعاء الدوال](#استدعاء-الدوال)
+  - [معلمات التوليد (v2.2.0)](#معلمات-التوليد-v220)
+    - [وضع التفكير ومستوى الاستدلال](#وضع-التفكير-ومستوى-الاستدلال)
+    - [تسلسلات الإيقاف (stop)](#تسلسلات-الإيقاف-stop)
+    - [أخذ العينات بالنواة (top_p)](#أخذ-العينات-بالنواة-top_p)
+    - [التحكم باختيار الأداة (tool_choice)](#التحكم-باختيار-الأداة-tool_choice)
+    - [الاحتمالات اللوغاريتمية (logprobs)](#الاحتمالات-اللوغاريتمية-logprobs)
+    - [معرّف المستخدم النهائي](#معرّف-المستخدم-النهائي)
+    - [حقل name على الرسائل](#حقل-name-على-الرسائل)
+    - [أدوات الوضع الصارم (strict)](#أدوات-الوضع-الصارم-strict)
   - [تكامل مع الأطر](#-تكامل-مع-الأطر)
 - [🆕 دليل الترحيل](#-دليل-الترحيل)
 - [📝 سجل التغييرات](#-سجل-التغييرات)
@@ -46,11 +55,19 @@
 - **تكامل API سلس**: واجهة تعتمد على PHP لميزات الذكاء الاصطناعي في DeepSeek.
 - **نمط الباني السلس**: أساليب قابلة للسلسلة لبناء الطلبات بطريقة بديهية.
 - **جاهز للمؤسسات**: تكامل مع عميل HTTP متوافق مع PSR-18.
-- **مرونة النماذج**: دعم لعدة نماذج من DeepSeek (Coder, Chat, وغيرها).
+- **أحدث نماذج DeepSeek V4**: دعم مباشر لـ `deepseek-v4-pro` و `deepseek-v4-flash` بنافذة سياق تصل إلى مليون رمز ووضعَي التفكير وعدم التفكير.
+- **تحكم كامل بمعلمات التوليد (v2.2.0)**: واجهات سلسلة لوضع التفكير، مستوى الاستدلال، تسلسلات الإيقاف، `top_p`، اختيار الأداة (`none` / `auto` / `required` / دالة مسماة)، `logprobs`، معرّف المستخدم النهائي، حقل `name` للرسائل، وأدوات الوضع الصارم.
 - **جاهز للبث**: دعم مدمج للتعامل مع الردود في الوقت الفعلي.
 - **العديد من عملاء HTTP**: يمكنك استخدام عميل `Guzzle http client` (افتراضي) أو `symfony http client` بسهولة.
 - **متوافق مع الأطر**: حزم Laravel و Symfony متاحة.
 
+> **النماذج المدعومة**
+>
+> - `Models::V4_PRO` — النموذج الرائد، أقصى عدد رموز للإخراج 384K.
+> - `Models::V4_FLASH` — سريع وموفّر، أقصى عدد رموز للإخراج 384K.
+>
+> النماذج القديمة `Models::CHAT`، `Models::CODER`، `Models::R1`، و `Models::R1Zero` مهملة وستتم إزالتها في v3.0.0. الأسماء البديلة `deepseek-chat` و `deepseek-reasoner` ستنسحب من DeepSeek API بتاريخ **2026-07-24**.
+
 ---
 
 ## 📦 التثبيت
@@ -83,8 +100,10 @@ echo $response;
 ```
 
 📌 الإعدادات الافتراضية المستخدمة:
-- النموذج: `deepseek-chat`
--  الحرارة: 0.8
+- النموذج: الافتراضي من API (لا يُرسل حقل `model` ما لم تستدعِ `withModel()`).
+- الحرارة: 1.3 (`TemperatureValues::GENERAL_CONVERSATION`).
+- أقصى عدد رموز: 4096.
+- صيغة الاستجابة: `text`.
 
 ### التكوين المتقدم
 
@@ -92,14 +111,14 @@ echo $response;
 use DeepSeek\DeepSeekClient;
 use DeepSeek\Enums\Models;
 
-$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com/v3', timeout:30, clientType:'guzzle');
+$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com', timeout:30, clientType:'guzzle');
 
 $response = $client
-    ->withModel(Models::CODER->value)
+    ->withModel(Models::V4_PRO->value)
     ->withStream()
-    ->withTemperature(1.2)
+    ->setTemperature(1.2)
     ->setMaxTokens(8192)
-    ->setResponseFormat('text')
+    ->setResponseFormat('text') // أو "json_object" بحذر.
     ->query('Explain quantum computing in simple terms')
     ->run();
 
@@ -150,7 +169,7 @@ echo 'API Response:'.$response;
 //  مع القيم الافتراضية للـ baseUrl و timeout
 $client = DeepSeekClient::build('your-api-key', clientType:'symfony')
 // مع التخصيص
-$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com/v3', timeout:30, clientType:'symfony');
+$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com', timeout:30, clientType:'symfony');
 
 $client->query('Explain quantum computing in simple terms')
        ->run();
@@ -165,7 +184,16 @@ $response = DeepSeekClient::build('your-api-key')
     ->getModelsList()
     ->run();
 
-echo $response; // {"object":"list","data":[{"id":"deepseek-chat","object":"model","owned_by":"deepseek"},{"id":"deepseek-reasoner","object":"model","owned_by":"deepseek"}]}
+echo $response;
+// {
+//   "object": "list",
+//   "data": [
+//     {"id": "deepseek-v4-pro",   "object": "model", "owned_by": "deepseek"},
+//     {"id": "deepseek-v4-flash", "object": "model", "owned_by": "deepseek"},
+//     {"id": "deepseek-chat",     "object": "model", "owned_by": "deepseek"},     // مهمل، ينسحب بتاريخ 2026-07-24
+//     {"id": "deepseek-reasoner", "object": "model", "owned_by": "deepseek"}      // مهمل، ينسحب بتاريخ 2026-07-24
+//   ]
+// }
 ```
 
 ###  استدعاء الدوال 
@@ -176,8 +204,121 @@ echo $response; // {"object":"list","data":[{"id":"deepseek-chat","object":"mode
 
 ---
 
-هل ترغب في أن أضع النسخ الثلاث (الإنجليزية + العربية + الصينية) ضمن ملف Markdown موحد؟
+### معلمات التوليد (v2.2.0)
+
+منذ الإصدار `v2.2.0` تُتيح الحزمة كامل واجهة معلمات التوليد عبر دوال سلسلة. **كل دالة جديدة اختيارية**: إذا لم تستدعِها، يبقى جسم الطلب مطابقًا حرفيًا لإصدار v2.1.x — لا يوجد ما يستوجب الترحيل.
+
+#### وضع التفكير ومستوى الاستدلال
+
+تدعم نماذج V4 خطوة "تفكير" مخصصة. استخدم [`setThinking()`](src/Traits/Client/HasGenerationParams.php) لتشغيلها أو إيقافها، و [`setReasoningEffort()`](src/Traits/Client/HasGenerationParams.php) لاختيار `high` (الافتراضي للطلبات العادية) أو `max` (موصى به لمسارات الوكلاء).
+
+```php
+use DeepSeek\DeepSeekClient;
+use DeepSeek\Enums\Configs\ReasoningEffort;
+use DeepSeek\Enums\Configs\ThinkingType;
+use DeepSeek\Enums\Models;
+
+$response = DeepSeekClient::build('your-api-key')
+    ->withModel(Models::V4_PRO->value)
+    ->setThinking(['type' => ThinkingType::ENABLED->value])
+    ->setReasoningEffort(ReasoningEffort::MAX->value)
+    ->query('Prove that there are infinitely many primes.')
+    ->run();
+```
+
+> ⚠️ في وضع التفكير يتجاهل DeepSeek API بصمت كلًّا من `temperature` و `top_p`، كما تُعيد `logprobs` / `top_logprobs` خطأ HTTP 400. انظر [وثائق نموذج الاستدلال](https://api-docs.deepseek.com/guides/reasoning_model).
+
+#### تسلسلات الإيقاف (stop)
+
+ما يصل إلى 16 تسلسلًا. مرّر سلسلة نصية واحدة أو مصفوفة؛ السلاسل المفردة تُحوَّل تلقائيًا إلى مصفوفة من عنصر واحد.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Write a haiku')
+    ->setStop(['###', "\n\n"])
+    ->run();
+```
+
+#### أخذ العينات بالنواة (top_p)
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Tell me a short story')
+    ->setTopP(0.95)
+    ->run();
+```
+
+#### التحكم باختيار الأداة (tool_choice)
+
+تقبل [`setToolChoice()`](src/Traits/Client/HasGenerationParams.php) القيم `"none"` و `"auto"` و `"required"` (إجبار استدعاء أداة)، أو شكل الدالة المسماة. وضع `"required"` المفقود سابقًا أصبح متاحًا الآن.
+
+```php
+use DeepSeek\Enums\Queries\ToolChoiceMode;
+
+// إجبار النموذج على استدعاء أي أداة
+$client->setTools($tools)
+       ->setToolChoice(ToolChoiceMode::REQUIRED->value);
+
+// إجبار النموذج على استدعاء دالة محددة بالاسم
+$client->setTools($tools)
+       ->setToolChoice([
+           'type' => 'function',
+           'function' => ['name' => 'get_weather'],
+       ]);
+```
+
+#### الاحتمالات اللوغاريتمية (logprobs)
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello')
+    ->setLogprobs(true)
+    ->setTopLogprobs(5)
+    ->run();
+```
+
+#### معرّف المستخدم النهائي
+
+لعزل حدود المعدل، السلامة، وعزل ذاكرة التخزين المؤقت (KV-cache). يُرسل على السلك باسم حقل `user` وفق مواصفة OpenAI.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setUserId('user-42')
+    ->query('Hello')
+    ->run();
+```
 
+#### حقل name على الرسائل
+
+معامل ثالث اختياري على `query()` (و `buildQuery()`) للتمييز بين المشاركين الذين يحملون نفس الدور وفق مواصفة OpenAI. الاستدعاءات القديمة بمعاملين لا تتأثر.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello, I am Alice.', 'user', 'alice')
+    ->query('Hello, I am Bob.',   'user', 'bob')
+    ->run();
+```
+
+#### أدوات الوضع الصارم (strict)
+
+تُضيف [`setStrictTool()`](src/Traits/Client/HasToolsFunctionCalling.php) دالة بأداة مع `strict: true`، مما يُلزم النموذج بإنتاج وسائط مطابقة تمامًا لمخطط JSON. تتكامل بأمان مع `setTools()` — تُلحِق ولا تستبدل.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setStrictTool(
+        name: 'get_weather',
+        parameters: [
+            'type' => 'object',
+            'properties' => ['city' => ['type' => 'string']],
+            'required' => ['city'],
+        ],
+        description: 'Get the current weather for a city.',
+    )
+    ->query('What is the weather like in Cairo?')
+    ->run();
+```
+
+---
 
 ### 🛠 تكامل مع الأطر
 
diff --git a/README-CN.md b/README-CN.md
index 0641316..78de4e7 100644
--- a/README-CN.md
+++ b/README-CN.md
@@ -32,6 +32,15 @@
   - [使用 Symfony HttpClient](#使用-symfony-httpclient)
   - [获取模型列表](#获取模型列表)
   - [函数调用](#函数调用)
+  - [生成参数（v2.2.0）](#生成参数v220)
+    - [思考模式与推理强度](#思考模式与推理强度)
+    - [停止序列（stop）](#停止序列stop)
+    - [核采样（top_p）](#核采样top_p)
+    - [工具选择控制（tool_choice）](#工具选择控制tool_choice)
+    - [对数概率（logprobs）](#对数概率logprobs)
+    - [最终用户标识](#最终用户标识)
+    - [消息 name 字段](#消息-name-字段)
+    - [严格模式工具（strict）](#严格模式工具strict)
   - [框架集成](#-框架集成)
 - [🆕 迁移指南](#-迁移指南)
 - [📝 更新日志](#-更新日志)
@@ -46,10 +55,18 @@
 - **无缝 API 集成**: DeepSeek AI 功能的 PHP 优先接口
 - **构建器模式**: 直观的链接请求构建方法
 - **企业级别**: 符合 PSR-18 规范
-- **模型灵活性**: 支持多种 DeepSeek 模型（Coder、Chat 等）
+- **最新 DeepSeek V4 模型**: 一流支持 `deepseek-v4-pro` 和 `deepseek-v4-flash`，具备 1M 令牌上下文窗口及思考 / 非思考模式
+- **完整的生成参数控制（v2.2.0）**: 提供思考模式、推理强度、停止序列、`top_p`、工具选择（`none` / `auto` / `required` / 命名函数）、`logprobs`、最终用户标识、消息 `name` 字段以及严格模式工具的链式 setter
 - **流式传输**: 内置对实时响应处理的支持
 - **框架友好**: 提供 Laravel 和 Symfony 包
 
+> **受支持的模型**
+>
+> - `Models::V4_PRO` — 旗舰模型，最大输出 384K 令牌。
+> - `Models::V4_FLASH` — 快速、经济，最大输出 384K 令牌。
+>
+> 旧版 `Models::CHAT`、`Models::CODER`、`Models::R1` 与 `Models::R1Zero` 已弃用，将在 v3.0.0 中移除。`deepseek-chat` 与 `deepseek-reasoner` 别名将于 **2026-07-24** 从 DeepSeek API 中下线。
+
 ---
 
 ## 📦 安装
@@ -82,23 +99,25 @@ echo $response;
 ```
 
 📌 默认配置:
-- Model: `deepseek-chat`
-- Temperature: 0.8
+- 模型: API 默认（除非调用 `withModel()`，否则不发送 `model` 字段）
+- 温度: 1.3（`TemperatureValues::GENERAL_CONVERSATION`）
+- 最大令牌数: 4096
+- 响应格式: `text`
 
-### Advanced Configuration
+### 高级配置
 
 ```php
 use DeepSeek\DeepSeekClient;
 use DeepSeek\Enums\Models;
 
-$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com/v3', timeout:30, clientType:'guzzle');
+$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com', timeout:30, clientType:'guzzle');
 
 $response = $client
-    ->withModel(Models::CODER->value)
+    ->withModel(Models::V4_PRO->value)
     ->withStream()
-    ->withTemperature(1.2)
+    ->setTemperature(1.2)
     ->setMaxTokens(8192)
-    ->setResponseFormat('text')
+    ->setResponseFormat('text') // 或 "json_object"，请谨慎使用。
     ->query('Explain quantum computing in simple terms')
     ->run();
 
@@ -139,16 +158,17 @@ echo 'API Response:'.$response;
 
 ---
 
-### Use with Symfony HttpClient
-the package already built with `symfony Http client`,  if you need to use package with `symfony` Http Client , it is easy to achieve that, just pass `clientType:'symfony'` with `build` function.
+### 使用 Symfony HttpClient
+
+本包已内置 `symfony Http client`。若需使用 Symfony 的 HTTP 客户端，只需在 `build` 函数中传入 `clientType:'symfony'` 即可。
 
-ex with symfony:
+Symfony 示例：
 
 ```php
-//  with defaults baseUrl and timeout
+// 使用默认的 baseUrl 和 timeout
 $client = DeepSeekClient::build('your-api-key', clientType:'symfony')
-// with customization
-$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com/v3', timeout:30, clientType:'symfony');
+// 自定义配置
+$client = DeepSeekClient::build(apiKey:'your-api-key', baseUrl:'https://api.deepseek.com', timeout:30, clientType:'symfony');
 
 $client->query('Explain quantum computing in simple terms')
        ->run();
@@ -163,7 +183,16 @@ $response = DeepSeekClient::build('your-api-key')
     ->getModelsList()
     ->run();
 
-echo $response; // {"object":"list","data":[{"id":"deepseek-chat","object":"model","owned_by":"deepseek"},{"id":"deepseek-reasoner","object":"model","owned_by":"deepseek"}]}
+echo $response;
+// {
+//   "object": "list",
+//   "data": [
+//     {"id": "deepseek-v4-pro",   "object": "model", "owned_by": "deepseek"},
+//     {"id": "deepseek-v4-flash", "object": "model", "owned_by": "deepseek"},
+//     {"id": "deepseek-chat",     "object": "model", "owned_by": "deepseek"},     // 已弃用，2026-07-24 下线
+//     {"id": "deepseek-reasoner", "object": "model", "owned_by": "deepseek"}      // 已弃用，2026-07-24 下线
+//   ]
+// }
 ```
 
 ### 函数调用
@@ -172,6 +201,123 @@ echo $response; // {"object":"list","data":[{"id":"deepseek-chat","object":"mode
 你可以在文档中查看有关函数调用的详细信息：
 [FUNCTION-CALLING.md](docs/FUNCTION-CALLING.md)
 
+---
+
+### 生成参数（v2.2.0）
+
+从 `v2.2.0` 起，本客户端通过链式 setter 暴露了完整的 DeepSeek 生成参数。**所有新 setter 都是可选的**：若不调用，请求体与 v2.1.x 完全一致 — 无需任何迁移。
+
+#### 思考模式与推理强度
+
+V4 模型支持专门的"思考"推理阶段。使用 [`setThinking()`](src/Traits/Client/HasGenerationParams.php) 切换开关，使用 [`setReasoningEffort()`](src/Traits/Client/HasGenerationParams.php) 在 `high`（普通请求的默认值）与 `max`（推荐用于 Agent 流程）之间选择。
+
+```php
+use DeepSeek\DeepSeekClient;
+use DeepSeek\Enums\Configs\ReasoningEffort;
+use DeepSeek\Enums\Configs\ThinkingType;
+use DeepSeek\Enums\Models;
+
+$response = DeepSeekClient::build('your-api-key')
+    ->withModel(Models::V4_PRO->value)
+    ->setThinking(['type' => ThinkingType::ENABLED->value])
+    ->setReasoningEffort(ReasoningEffort::MAX->value)
+    ->query('Prove that there are infinitely many primes.')
+    ->run();
+```
+
+> ⚠️ 在思考模式下，DeepSeek API 会静默忽略 `temperature` / `top_p`，并对 `logprobs` / `top_logprobs` 返回 HTTP 400。请参考 [推理模型文档](https://api-docs.deepseek.com/guides/reasoning_model)。
+
+#### 停止序列（stop）
+
+最多 16 个停止序列。可传入字符串或数组；单个字符串会被规范化为单元素数组。
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Write a haiku')
+    ->setStop(['###', "\n\n"])
+    ->run();
+```
+
+#### 核采样（top_p）
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Tell me a short story')
+    ->setTopP(0.95)
+    ->run();
+```
+
+#### 工具选择控制（tool_choice）
+
+[`setToolChoice()`](src/Traits/Client/HasGenerationParams.php) 接受 `"none"`、`"auto"`、`"required"`（强制调用工具）或命名函数数组形式。之前缺失的 `"required"` 模式现在已可用。
+
+```php
+use DeepSeek\Enums\Queries\ToolChoiceMode;
+
+// 强制模型调用任意工具
+$client->setTools($tools)
+       ->setToolChoice(ToolChoiceMode::REQUIRED->value);
+
+// 强制模型调用指定名称的函数
+$client->setTools($tools)
+       ->setToolChoice([
+           'type' => 'function',
+           'function' => ['name' => 'get_weather'],
+       ]);
+```
+
+#### 对数概率（logprobs）
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello')
+    ->setLogprobs(true)
+    ->setTopLogprobs(5)
+    ->run();
+```
+
+#### 最终用户标识
+
+用于速率限制隔离、内容安全和 KV 缓存隔离。在请求中以 OpenAI 规范的 `user` 字段发送。
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setUserId('user-42')
+    ->query('Hello')
+    ->run();
+```
+
+#### 消息 name 字段
+
+`query()`（与 `buildQuery()`）的可选第 3 个参数，遵循 OpenAI 规范用于区分同一角色下的不同参与者。原有的两参数调用保持不变。
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello, I am Alice.', 'user', 'alice')
+    ->query('Hello, I am Bob.',   'user', 'bob')
+    ->run();
+```
+
+#### 严格模式工具（strict）
+
+[`setStrictTool()`](src/Traits/Client/HasToolsFunctionCalling.php) 追加一个带 `strict: true` 的函数工具，强制模型生成完全符合 JSON Schema 的参数。可与 `setTools()` 安全组合 — 它是追加而非替换。
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setStrictTool(
+        name: 'get_weather',
+        parameters: [
+            'type' => 'object',
+            'properties' => ['city' => ['type' => 'string']],
+            'required' => ['city'],
+        ],
+        description: 'Get the current weather for a city.',
+    )
+    ->query('What is the weather like in Cairo?')
+    ->run();
+```
+
+---
 
 ### 🛠 框架集成
 
diff --git a/README.md b/README.md
index 8900581..acd29fa 100644
--- a/README.md
+++ b/README.md
@@ -32,6 +32,15 @@
   - [Use with Symfony HttpClient](#use-with-symfony-httpclient)
   - [Get Models List](#get-models-list)
   - [Function Calling](#function-calling)
+  - [Generation Parameters (v2.2.0)](#generation-parameters-v220)
+    - [Thinking mode and reasoning effort](#thinking-mode-and-reasoning-effort)
+    - [Stop sequences](#stop-sequences)
+    - [Nucleus sampling (top_p)](#nucleus-sampling-top_p)
+    - [Tool choice control](#tool-choice-control)
+    - [Log probabilities](#log-probabilities)
+    - [End-user identifier](#end-user-identifier)
+    - [Message name field](#message-name-field)
+    - [Strict mode tools](#strict-mode-tools)
   - [Framework Integration](#-framework-integration)
 - [🆕 Migration Guide](#-migration-guide)
 - [📝 Changelog](#-changelog)
@@ -47,6 +56,7 @@
 - **Fluent Builder Pattern**: Chainable methods for intuitive request building.
 - **Enterprise Ready**: PSR-18 compliant HTTP client integration.
 - **Latest DeepSeek V4 Models**: First-class support for `deepseek-v4-pro` and `deepseek-v4-flash` with 1M-token context windows and thinking / non-thinking modes.
+- **Full Generation Parameter Control (v2.2.0)**: Fluent setters for thinking mode, reasoning effort, stop sequences, `top_p`, tool choice (`none` / `auto` / `required` / named function), `logprobs`, end-user identifier, message `name` field, and strict-mode tools.
 - **Streaming Ready**: Built-in support for real-time response handling.
 - **Many Http Clients**: easy to use `Guzzle http client` (default) , or `symfony http client`.
 - **Framework Friendly**: Laravel & Symfony packages available.
@@ -192,6 +202,123 @@ Function Calling allows the model to call external tools to enhance its capabili
 
 You Can check the documentation for function calling in [FUNCTION-CALLING.md](docs/FUNCTION-CALLING.md)
 
+---
+
+### Generation Parameters (v2.2.0)
+
+Since `v2.2.0` the client exposes the full DeepSeek generation surface as fluent setters. **Every new setter is opt-in**: when you don't call it, the request body is byte-identical to v2.1.x — there's nothing to migrate.
+
+#### Thinking mode and reasoning effort
+
+V4 models support a dedicated "thinking" reasoning step. Use [`setThinking()`](src/Traits/Client/HasGenerationParams.php) to toggle it on or off, and [`setReasoningEffort()`](src/Traits/Client/HasGenerationParams.php) to choose between `high` (default for normal requests) and `max` (recommended for agent flows).
+
+```php
+use DeepSeek\DeepSeekClient;
+use DeepSeek\Enums\Configs\ReasoningEffort;
+use DeepSeek\Enums\Configs\ThinkingType;
+use DeepSeek\Enums\Models;
+
+$response = DeepSeekClient::build('your-api-key')
+    ->withModel(Models::V4_PRO->value)
+    ->setThinking(['type' => ThinkingType::ENABLED->value])
+    ->setReasoningEffort(ReasoningEffort::MAX->value)
+    ->query('Prove that there are infinitely many primes.')
+    ->run();
+```
+
+> ⚠️ In thinking mode the DeepSeek API silently ignores `temperature` / `top_p`, and `logprobs` / `top_logprobs` return HTTP 400. See the [reasoning model docs](https://api-docs.deepseek.com/guides/reasoning_model).
+
+#### Stop sequences
+
+Up to 16 stop sequences. Pass a single string or an array; single strings are normalized to a one-element array.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Write a haiku')
+    ->setStop(['###', "\n\n"])
+    ->run();
+```
+
+#### Nucleus sampling (top_p)
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Tell me a short story')
+    ->setTopP(0.95)
+    ->run();
+```
+
+#### Tool choice control
+
+[`setToolChoice()`](src/Traits/Client/HasGenerationParams.php) accepts `"none"`, `"auto"`, `"required"` (force a tool call), or the named-function array shape. The previously missing `"required"` mode is now reachable.
+
+```php
+use DeepSeek\Enums\Queries\ToolChoiceMode;
+
+// Force the model to call any tool
+$client->setTools($tools)
+       ->setToolChoice(ToolChoiceMode::REQUIRED->value);
+
+// Force a specific named function
+$client->setTools($tools)
+       ->setToolChoice([
+           'type' => 'function',
+           'function' => ['name' => 'get_weather'],
+       ]);
+```
+
+#### Log probabilities
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello')
+    ->setLogprobs(true)
+    ->setTopLogprobs(5)
+    ->run();
+```
+
+#### End-user identifier
+
+For rate-limit isolation, content safety, and KV-cache isolation. Sent on the wire as the OpenAI-spec `user` field.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setUserId('user-42')
+    ->query('Hello')
+    ->run();
+```
+
+#### Message `name` field
+
+Optional third parameter on `query()` (and `buildQuery()`) to differentiate participants of the same role per OpenAI spec. Existing 2-argument calls are unchanged.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->query('Hello, I am Alice.', 'user', 'alice')
+    ->query('Hello, I am Bob.',   'user', 'bob')
+    ->run();
+```
+
+#### Strict mode tools
+
+[`setStrictTool()`](src/Traits/Client/HasToolsFunctionCalling.php) appends a function tool with `strict: true`, instructing the model to produce arguments that conform exactly to the JSON schema. Composes safely with `setTools()` — it appends, never replaces.
+
+```php
+$response = DeepSeekClient::build('your-api-key')
+    ->setStrictTool(
+        name: 'get_weather',
+        parameters: [
+            'type' => 'object',
+            'properties' => ['city' => ['type' => 'string']],
+            'required' => ['city'],
+        ],
+        description: 'Get the current weather for a city.',
+    )
+    ->query('What is the weather like in Cairo?')
+    ->run();
+```
+
+---
 
 ### 🛠 Framework Integration
 
diff --git a/composer.json b/composer.json
index 626c9e4..e354fdb 100644
--- a/composer.json
+++ b/composer.json
@@ -49,7 +49,7 @@
             "role": "creator"
         }
     ],
-    "version": "2.1.0",
+    "version": "2.2.0",
     "require": {
         "php": "^8.2.0",
         "nyholm/psr7": "^1.8",
diff --git a/docs/FUNCTION-CALLING.md b/docs/FUNCTION-CALLING.md
index 268ac08..061a229 100644
--- a/docs/FUNCTION-CALLING.md
+++ b/docs/FUNCTION-CALLING.md
@@ -196,4 +196,69 @@ Output response like :-
 
 When using V4 models with thinking mode enabled (or the legacy `DeepSeek-R1`), assistant responses include a `reasoning_content` field at the same level as `content`. **This field MUST be echoed back on the next tool turn**; otherwise the DeepSeek API returns HTTP 400.
 
-See the [DeepSeek reasoning model docs](https://api-docs.deepseek.com/guides/reasoning_model) for details. Helpers for reading `reasoning_content` off the response and passing it back into the next request will land in `v2.2.0` together with `setThinking()` / `setReasoningEffort()`.
+Since `v2.2.0` you can toggle thinking mode and reasoning effort directly:
+
+```php
+use DeepSeek\Enums\Configs\ReasoningEffort;
+use DeepSeek\Enums\Configs\ThinkingType;
+
+$client
+    ->setThinking(['type' => ThinkingType::ENABLED->value])
+    ->setReasoningEffort(ReasoningEffort::MAX->value);
+```
+
+The matching response-side helper for extracting `reasoning_content` and re-injecting it on the next turn is on the `v2.2.x` roadmap (see [CHANGELOG.md](../CHANGELOG.md)). In the meantime, callers should decode the response JSON and pass `reasoning_content` back manually inside the assistant message.
+
+See the [DeepSeek reasoning model docs](https://api-docs.deepseek.com/guides/reasoning_model) for the full caveat list — notably that `temperature` / `top_p` are silently ignored in thinking mode and `logprobs` / `top_logprobs` return HTTP 400.
+
+---
+
+### Tool choice control (v2.2.0)
+
+By default DeepSeek decides freely whether to call a tool. Use [`setToolChoice()`](../src/Traits/Client/HasGenerationParams.php) to constrain that behavior:
+
+| Mode | Effect |
+|---|---|
+| `'none'` | The model will not call any tool. |
+| `'auto'` | The model decides (default behavior). |
+| `'required'` | The model MUST call at least one tool. |
+| `['type' => 'function', 'function' => ['name' => 'foo']]` | The model MUST call the named function. |
+
+```php
+use DeepSeek\Enums\Queries\ToolChoiceMode;
+
+// Force the model to call a tool (any tool)
+$client->setTools($tools)
+       ->setToolChoice(ToolChoiceMode::REQUIRED->value);
+
+// Force the model to call get_weather specifically
+$client->setTools($tools)
+       ->setToolChoice([
+           'type' => 'function',
+           'function' => ['name' => 'get_weather'],
+       ]);
+```
+
+---
+
+### Strict mode tools (v2.2.0)
+
+`strict: true` on a function definition guarantees the model produces arguments that conform exactly to the JSON schema. Use [`setStrictTool()`](../src/Traits/Client/HasToolsFunctionCalling.php) instead of building the array yourself:
+
+```php
+$client
+    ->setStrictTool(
+        name: 'get_weather',
+        parameters: [
+            'type' => 'object',
+            'properties' => [
+                'city' => ['type' => 'string', 'description' => 'The city name'],
+            ],
+            'required' => ['city'],
+        ],
+        description: 'Get the current weather in a given city',
+    )
+    ->query('What is the weather like in Cairo?');
+```
+
+`setStrictTool()` **appends** to whatever `setTools()` already set; it never replaces. You can mix strict and non-strict tools in the same request by chaining calls.
diff --git a/src/DeepSeekClient.php b/src/DeepSeekClient.php
index 3d55b4e..1dbedf3 100644
--- a/src/DeepSeekClient.php
+++ b/src/DeepSeekClient.php
@@ -11,6 +11,7 @@
 use DeepSeek\Enums\Requests\QueryFlags;
 use DeepSeek\Factories\ApiFactory;
 use DeepSeek\Resources\Resource;
+use DeepSeek\Traits\Client\HasGenerationParams;
 use DeepSeek\Traits\Client\HasToolsFunctionCalling;
 use DeepSeek\Traits\Resources\HasChat;
 use DeepSeek\Traits\Resources\HasCoder;
@@ -19,6 +20,7 @@
 class DeepSeekClient implements ClientContract
 {
     use HasChat, HasCoder;
+    use HasGenerationParams;
     use HasToolsFunctionCalling;
 
     /**
@@ -93,6 +95,12 @@ public function run(): string
             ],
         ];
 
+        foreach ($this->getOptionalRequestParams() as $key => $value) {
+            if ($value !== null) {
+                $requestData[$key] = $value;
+            }
+        }
+
         $this->setResult((new Resource($this->httpClient, $this->endpointSuffixes))->sendRequest($requestData, $this->requestMethod));
 
         return $this->getResult()->getContent();
@@ -122,11 +130,14 @@ public static function build(string $apiKey, ?string $baseUrl = null, ?int $time
     /**
      * Add a query to the accumulated queries list.
      *
+     * @param  string|null  $name  Optional OpenAI-spec "name" field on the message.
+     *                             When non-null it differentiates participants of the
+     *                             same role. Omitted from the message when null.
      * @return self The current instance for method chaining.
      */
-    public function query(string $content, ?string $role = 'user'): self
+    public function query(string $content, ?string $role = 'user', ?string $name = null): self
     {
-        $this->queries[] = $this->buildQuery($content, $role);
+        $this->queries[] = $this->buildQuery($content, $role, $name);
 
         return $this;
     }
@@ -203,12 +214,18 @@ public function setResponseFormat(string $type): self
         return $this;
     }
 
-    public function buildQuery(string $content, ?string $role = null): array
+    public function buildQuery(string $content, ?string $role = null, ?string $name = null): array
     {
-        return [
+        $query = [
             'role' => $role ?: QueryRoles::USER->value,
             'content' => $content,
         ];
+
+        if ($name !== null) {
+            $query['name'] = $name;
+        }
+
+        return $query;
     }
 
     /**
diff --git a/src/Enums/Configs/ReasoningEffort.php b/src/Enums/Configs/ReasoningEffort.php
new file mode 100644
index 0000000..4e09808
--- /dev/null
+++ b/src/Enums/Configs/ReasoningEffort.php
@@ -0,0 +1,17 @@
+<?php
+
+namespace DeepSeek\Enums\Configs;
+
+/**
+ * Reasoning effort levels for thinking-capable models (DeepSeek V4).
+ *
+ * Per the DeepSeek reasoning model docs, "high" is the default for normal
+ * requests and "max" is recommended for agent flows (e.g. Claude Code / OpenCode).
+ *
+ * @see https://api-docs.deepseek.com/guides/reasoning_model
+ */
+enum ReasoningEffort: string
+{
+    case HIGH = 'high';
+    case MAX = 'max';
+}
diff --git a/src/Enums/Configs/ThinkingType.php b/src/Enums/Configs/ThinkingType.php
new file mode 100644
index 0000000..6a1918e
--- /dev/null
+++ b/src/Enums/Configs/ThinkingType.php
@@ -0,0 +1,17 @@
+<?php
+
+namespace DeepSeek\Enums\Configs;
+
+/**
+ * Thinking mode toggle for DeepSeek V4 models.
+ *
+ * The API field "thinking.type" defaults to "enabled" on the server side.
+ * Set "disabled" to bypass the reasoning step.
+ *
+ * @see https://api-docs.deepseek.com/guides/reasoning_model
+ */
+enum ThinkingType: string
+{
+    case ENABLED = 'enabled';
+    case DISABLED = 'disabled';
+}
diff --git a/src/Enums/Queries/ToolChoiceMode.php b/src/Enums/Queries/ToolChoiceMode.php
new file mode 100644
index 0000000..48c2eb5
--- /dev/null
+++ b/src/Enums/Queries/ToolChoiceMode.php
@@ -0,0 +1,18 @@
+<?php
+
+namespace DeepSeek\Enums\Queries;
+
+/**
+ * Pre-defined string modes for the OpenAI-style "tool_choice" parameter.
+ *
+ * Named tool choice (forcing a specific function) is expressed as an array:
+ *   ['type' => 'function', 'function' => ['name' => 'my_function']]
+ *
+ * @see https://api-docs.deepseek.com/api/create-chat-completion
+ */
+enum ToolChoiceMode: string
+{
+    case NONE = 'none';
+    case AUTO = 'auto';
+    case REQUIRED = 'required';
+}
diff --git a/src/Enums/Requests/QueryFlags.php b/src/Enums/Requests/QueryFlags.php
index 29e2586..f421fe3 100644
--- a/src/Enums/Requests/QueryFlags.php
+++ b/src/Enums/Requests/QueryFlags.php
@@ -11,4 +11,60 @@ enum QueryFlags: string
     case MAX_TOKENS = 'max_tokens';
     case TOOLS = 'tools';
     case RESPONSE_FORMAT = 'response_format';
+
+    /**
+     * Up to 16 stop sequences (string or string[]).
+     *
+     * @see https://api-docs.deepseek.com/api/create-chat-completion
+     */
+    case STOP = 'stop';
+
+    /**
+     * Nucleus sampling parameter (0..1). Alternative to temperature.
+     *
+     * @see https://api-docs.deepseek.com/api/create-chat-completion
+     */
+    case TOP_P = 'top_p';
+
+    /**
+     * Tool choice control: "none" | "auto" | "required" | {"type": "function", "function": {"name": "..."}}.
+     *
+     * @see https://api-docs.deepseek.com/api/create-chat-completion
+     */
+    case TOOL_CHOICE = 'tool_choice';
+
+    /**
+     * Whether to return log probabilities of the output tokens.
+     *
+     * @see https://api-docs.deepseek.com/api/create-chat-completion
+     */
+    case LOGPROBS = 'logprobs';
+
+    /**
+     * Number of most likely tokens to return at each token position with log probabilities.
+     *
+     * @see https://api-docs.deepseek.com/api/create-chat-completion
+     */
+    case TOP_LOGPROBS = 'top_logprobs';
+
+    /**
+     * End-user identifier for rate-limit isolation, content safety, and KV-cache isolation.
+     *
+     * @see https://api-docs.deepseek.com/quick_start/rate_limit
+     */
+    case USER = 'user';
+
+    /**
+     * Thinking mode configuration: ["type" => "enabled" | "disabled"].
+     *
+     * @see https://api-docs.deepseek.com/guides/reasoning_model
+     */
+    case THINKING = 'thinking';
+
+    /**
+     * Reasoning effort: "high" | "max".
+     *
+     * @see https://api-docs.deepseek.com/guides/reasoning_model
+     */
+    case REASONING_EFFORT = 'reasoning_effort';
 }
diff --git a/src/Traits/Client/HasGenerationParams.php b/src/Traits/Client/HasGenerationParams.php
new file mode 100644
index 0000000..f1e66cd
--- /dev/null
+++ b/src/Traits/Client/HasGenerationParams.php
@@ -0,0 +1,185 @@
+<?php
+
+namespace DeepSeek\Traits\Client;
+
+use DeepSeek\Enums\Requests\QueryFlags;
+
+/**
+ * Holds the v2.2.0 additive generation parameter setters and their backing
+ * nullable state. Every property defaults to null and is omitted from the
+ * outgoing request body unless its corresponding setter has been called.
+ *
+ * This is the backward-compatibility contract: existing v2.1.x callers
+ * that never invoke any of these setters will see a byte-identical request
+ * payload compared to the previous release.
+ */
+trait HasGenerationParams
+{
+    /**
+     * Up to 16 stop sequences. Always stored as an array of strings.
+     */
+    protected ?array $stop = null;
+
+    /**
+     * Nucleus sampling parameter in the [0, 1] range.
+     */
+    protected ?float $topP = null;
+
+    /**
+     * OpenAI-style tool_choice value. Either one of the string modes
+     * ("none" | "auto" | "required") or a named function array shape.
+     *
+     * @var string|array<string,mixed>|null
+     */
+    protected string|array|null $toolChoice = null;
+
+    /**
+     * Whether the API should return log probabilities for output tokens.
+     */
+    protected ?bool $logprobs = null;
+
+    /**
+     * Number of most-likely tokens to include in the logprobs response.
+     */
+    protected ?int $topLogprobs = null;
+
+    /**
+     * End-user identifier. Sent as the OpenAI-spec "user" field.
+     */
+    protected ?string $userId = null;
+
+    /**
+     * Thinking-mode configuration array, e.g. ["type" => "enabled"].
+     *
+     * @var array<string,mixed>|null
+     */
+    protected ?array $thinking = null;
+
+    /**
+     * Reasoning effort level ("high" | "max").
+     */
+    protected ?string $reasoningEffort = null;
+
+    /**
+     * Set the stop sequences. Accepts a single string or an array of strings.
+     *
+     * @param  string|array<int,string>  $stop  Up to 16 stop sequences.
+     * @return self The current instance for method chaining.
+     */
+    public function setStop(string|array $stop): self
+    {
+        $this->stop = is_string($stop) ? [$stop] : array_values($stop);
+
+        return $this;
+    }
+
+    /**
+     * Set the top_p nucleus-sampling parameter.
+     *
+     * @return self The current instance for method chaining.
+     */
+    public function setTopP(float $topP): self
+    {
+        $this->topP = $topP;
+
+        return $this;
+    }
+
+    /**
+     * Set the OpenAI-style tool_choice value.
+     *
+     * @param  string|array<string,mixed>  $toolChoice  "none" | "auto" | "required"
+     *                                                  or ["type" => "function", "function" => ["name" => "..."]].
+     * @return self The current instance for method chaining.
+     */
+    public function setToolChoice(string|array $toolChoice): self
+    {
+        $this->toolChoice = $toolChoice;
+
+        return $this;
+    }
+
+    /**
+     * Enable or disable token log probabilities in the API response.
+     *
+     * @return self The current instance for method chaining.
+     */
+    public function setLogprobs(bool $enabled): self
+    {
+        $this->logprobs = $enabled;
+
+        return $this;
+    }
+
+    /**
+     * Set the number of most-likely tokens to include with log probabilities.
+     *
+     * @return self The current instance for method chaining.
+     */
+    public function setTopLogprobs(int $count): self
+    {
+        $this->topLogprobs = $count;
+
+        return $this;
+    }
+
+    /**
+     * Set the end-user identifier sent as the OpenAI-spec "user" field.
+     *
+     * @return self The current instance for method chaining.
+     */
+    public function setUserId(string $userId): self
+    {
+        $this->userId = $userId;
+
+        return $this;
+    }
+
+    /**
+     * Set the thinking-mode configuration array.
+     *
+     * @param  array<string,mixed>  $config  Typically ["type" => "enabled"] or ["type" => "disabled"].
+     * @return self The current instance for method chaining.
+     */
+    public function setThinking(array $config): self
+    {
+        $this->thinking = $config;
+
+        return $this;
+    }
+
+    /**
+     * Set the reasoning effort level ("high" | "max").
+     *
+     * @return self The current instance for method chaining.
+     */
+    public function setReasoningEffort(string $effort): self
+    {
+        $this->reasoningEffort = $effort;
+
+        return $this;
+    }
+
+    /**
+     * Return the optional v2.2.0 parameters keyed by API field name.
+     *
+     * Callers MUST filter out null values before merging into the request
+     * body — that is what preserves byte-identical request payloads for
+     * users who have not opted into any new setter.
+     *
+     * @return array<string,mixed>
+     */
+    protected function getOptionalRequestParams(): array
+    {
+        return [
+            QueryFlags::STOP->value => $this->stop,
+            QueryFlags::TOP_P->value => $this->topP,
+            QueryFlags::TOOL_CHOICE->value => $this->toolChoice,
+            QueryFlags::LOGPROBS->value => $this->logprobs,
+            QueryFlags::TOP_LOGPROBS->value => $this->topLogprobs,
+            QueryFlags::USER->value => $this->userId,
+            QueryFlags::THINKING->value => $this->thinking,
+            QueryFlags::REASONING_EFFORT->value => $this->reasoningEffort,
+        ];
+    }
+}
diff --git a/src/Traits/Client/HasToolsFunctionCalling.php b/src/Traits/Client/HasToolsFunctionCalling.php
index 9cae45b..d06b950 100644
--- a/src/Traits/Client/HasToolsFunctionCalling.php
+++ b/src/Traits/Client/HasToolsFunctionCalling.php
@@ -17,6 +17,43 @@ public function setTools(array $tools): self
         return $this;
     }
 
+    /**
+     * Append a single function tool with structured-output strict mode enabled.
+     *
+     * Strict mode is a DeepSeek Beta feature: when enabled the model is
+     * guaranteed to produce tool-call arguments that conform exactly to the
+     * provided JSON schema. See the DeepSeek API reference for caveats.
+     *
+     * This helper is purely additive — it appends to whatever the existing
+     * setTools() / prior setStrictTool() calls have already stored, so it
+     * never silently replaces an existing tool list.
+     *
+     * @param  string  $name  Function name.
+     * @param  array  $parameters  JSON-schema object describing the function arguments.
+     * @param  string|null  $description  Optional human-readable description.
+     * @return self The current instance for method chaining.
+     */
+    public function setStrictTool(string $name, array $parameters, ?string $description = null): self
+    {
+        $this->tools = $this->tools ?? [];
+
+        $function = ['name' => $name];
+
+        if ($description !== null) {
+            $function['description'] = $description;
+        }
+
+        $function['parameters'] = $parameters;
+        $function['strict'] = true;
+
+        $this->tools[] = [
+            'type' => 'function',
+            'function' => $function,
+        ];
+
+        return $this;
+    }
+
     /**
      * Add a query tool calls to the accumulated queries list.
      *
diff --git a/src/Traits/Resources/HasChat.php b/src/Traits/Resources/HasChat.php
index b03063e..2a83fb8 100644
--- a/src/Traits/Resources/HasChat.php
+++ b/src/Traits/Resources/HasChat.php
@@ -27,6 +27,13 @@ public function chat(): string
                 'type' => $this->responseFormatType,
             ],
         ];
+
+        foreach ($this->getOptionalRequestParams() as $key => $value) {
+            if ($value !== null) {
+                $requestData[$key] = $value;
+            }
+        }
+
         $this->queries = [];
         $this->setResult((new Chat($this->httpClient))->sendRequest($requestData));
 
diff --git a/src/Traits/Resources/HasCoder.php b/src/Traits/Resources/HasCoder.php
index 4ef5d52..e8a8df8 100644
--- a/src/Traits/Resources/HasCoder.php
+++ b/src/Traits/Resources/HasCoder.php
@@ -27,6 +27,13 @@ public function code(): string
                 'type' => $this->responseFormatType,
             ],
         ];
+
+        foreach ($this->getOptionalRequestParams() as $key => $value) {
+            if ($value !== null) {
+                $requestData[$key] = $value;
+            }
+        }
+
         $this->queries = [];
         $this->setResult((new Coder($this->httpClient))->sendRequest($requestData));
 
diff --git a/tests/Feature/V220ChangesTest.php b/tests/Feature/V220ChangesTest.php
new file mode 100644
index 0000000..9aeb8ba
--- /dev/null
+++ b/tests/Feature/V220ChangesTest.php
@@ -0,0 +1,469 @@
+<?php
+
+namespace Tests\Feature;
+
+use DeepSeek\DeepSeekClient;
+use DeepSeek\Enums\Configs\ReasoningEffort;
+use DeepSeek\Enums\Configs\ThinkingType;
+use DeepSeek\Enums\Models;
+use DeepSeek\Enums\Queries\ToolChoiceMode;
+use DeepSeek\Enums\Requests\QueryFlags;
+use Mockery;
+use Mockery\LegacyMockInterface;
+use Mockery\MockInterface;
+use Nyholm\Psr7\Factory\Psr17Factory;
+use Psr\Http\Client\ClientInterface;
+use Psr\Http\Message\RequestInterface;
+use Psr\Http\Message\ResponseInterface;
+
+afterEach(function () {
+    Mockery::close();
+});
+
+/**
+ * Boot a mocked PSR-18 client that captures the outgoing request body string
+ * into the supplied reference, and returns a minimal 200 OK JSON body.
+ */
+function v220MockClient(Psr17Factory $factory, ?string &$capturedBody): ClientInterface
+{
+    /** @var ClientInterface&LegacyMockInterface&MockInterface $httpClient */
+    $httpClient = Mockery::mock(ClientInterface::class);
+    $httpClient->shouldReceive('sendRequest')
+        ->once()
+        ->andReturnUsing(function (RequestInterface $request) use ($factory, &$capturedBody): ResponseInterface {
+            $capturedBody = (string) $request->getBody();
+
+            return $factory->createResponse(200)->withBody($factory->createStream('{"id":"ok"}'));
+        });
+
+    return $httpClient;
+}
+
+// =====================================================================
+// Backward-compatibility guards (these MUST pass first)
+// =====================================================================
+
+test('run() request body shape is byte-identical when no new setters called', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->withModel(Models::V4_FLASH->value)
+        ->query('Hello')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toBeArray()
+        ->and(array_keys($decoded))
+        ->toBe(['messages', 'model', 'stream', 'temperature', 'max_tokens', 'tools', 'response_format']);
+});
+
+test('chat() shortcut request body shape is byte-identical when no new setters called', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->withModel(Models::V4_FLASH->value)
+        ->query('Hello')
+        ->chat();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toBeArray()
+        ->and(array_keys($decoded))
+        ->toBe(['messages', 'model', 'stream', 'temperature', 'max_tokens', 'tools', 'response_format']);
+});
+
+test('code() shortcut request body shape is byte-identical when no new setters called', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->withModel(Models::V4_PRO->value)
+        ->query('def fib(n):')
+        ->code();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toBeArray()
+        ->and(array_keys($decoded))
+        ->toBe(['messages', 'model', 'stream', 'temperature', 'max_tokens', 'tools', 'response_format']);
+});
+
+// =====================================================================
+// Per-setter additive tests
+// =====================================================================
+
+test('setStop(string) adds stop as a single-element array', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setStop('###')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toHaveKey('stop')
+        ->and($decoded['stop'])->toBe(['###']);
+});
+
+test('setStop(array) adds stop as passed', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setStop(['###', '<END>', "\n\n"])
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['stop'])->toBe(['###', '<END>', "\n\n"]);
+});
+
+test('setTopP adds top_p to request body', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setTopP(0.95)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['top_p'])->toBe(0.95);
+});
+
+test('setToolChoice with string "auto" adds tool_choice', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setToolChoice('auto')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tool_choice'])->toBe('auto');
+});
+
+test('setToolChoice with named function array adds tool_choice', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    $named = ['type' => 'function', 'function' => ['name' => 'get_weather']];
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setToolChoice($named)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tool_choice'])->toBe($named);
+});
+
+test('setToolChoice with "required" reaches the wire', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setToolChoice(ToolChoiceMode::REQUIRED->value)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tool_choice'])->toBe('required');
+});
+
+test('setLogprobs(true) adds logprobs to request body', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setLogprobs(true)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toHaveKey('logprobs')
+        ->and($decoded['logprobs'])->toBeTrue();
+});
+
+test('setTopLogprobs adds top_logprobs to request body', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setTopLogprobs(5)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['top_logprobs'])->toBe(5);
+});
+
+test('setUserId adds OpenAI-spec user field', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setUserId('user-42')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['user'])->toBe('user-42');
+});
+
+test('setThinking adds thinking config to request body', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setThinking(['type' => ThinkingType::ENABLED->value])
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['thinking'])->toBe(['type' => 'enabled']);
+});
+
+test('setReasoningEffort adds reasoning_effort to request body', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setReasoningEffort(ReasoningEffort::MAX->value)
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['reasoning_effort'])->toBe('max');
+});
+
+test('all new setters combined produce all new keys plus all existing keys', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->withModel(Models::V4_PRO->value)
+        ->query('Hello')
+        ->setStop(['###'])
+        ->setTopP(0.9)
+        ->setToolChoice('auto')
+        ->setLogprobs(true)
+        ->setTopLogprobs(3)
+        ->setUserId('user-42')
+        ->setThinking(['type' => 'enabled'])
+        ->setReasoningEffort('high')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded)
+        ->toHaveKeys([
+            'messages', 'model', 'stream', 'temperature', 'max_tokens', 'tools', 'response_format',
+            'stop', 'top_p', 'tool_choice', 'logprobs', 'top_logprobs', 'user', 'thinking', 'reasoning_effort',
+        ])
+        ->and($decoded['model'])->toBe('deepseek-v4-pro');
+});
+
+// =====================================================================
+// name field on messages (#10)
+// =====================================================================
+
+test('query() with no name argument produces a message with only role and content', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('hi')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['messages'])->toBeArray()->toHaveCount(1)
+        ->and(array_keys($decoded['messages'][0]))->toBe(['role', 'content'])
+        ->and($decoded['messages'][0])->toBe(['role' => 'user', 'content' => 'hi']);
+});
+
+test('query() with name argument emits the name field', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('hi', 'user', 'alice')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['messages'][0])
+        ->toBe(['role' => 'user', 'content' => 'hi', 'name' => 'alice']);
+});
+
+test('buildQuery() with name argument returns the name key', function () {
+    $client = new DeepSeekClient(Mockery::mock(ClientInterface::class));
+
+    $result = $client->buildQuery('hi', null, 'bob');
+
+    expect($result)
+        ->toBe(['role' => 'user', 'content' => 'hi', 'name' => 'bob']);
+});
+
+// =====================================================================
+// Tool strict mode helper (#13)
+// =====================================================================
+
+test('setStrictTool adds a function tool with strict=true', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    $schema = [
+        'type' => 'object',
+        'properties' => ['city' => ['type' => 'string']],
+        'required' => ['city'],
+    ];
+
+    (new DeepSeekClient($httpClient))
+        ->query('Weather?')
+        ->setStrictTool('get_weather', $schema, 'Get the current weather for a city.')
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tools'])->toBeArray()->toHaveCount(1)
+        ->and($decoded['tools'][0])
+        ->toBe([
+            'type' => 'function',
+            'function' => [
+                'name' => 'get_weather',
+                'description' => 'Get the current weather for a city.',
+                'parameters' => $schema,
+                'strict' => true,
+            ],
+        ]);
+});
+
+test('setStrictTool called twice appends both tools', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setStrictTool('foo', ['type' => 'object'])
+        ->setStrictTool('bar', ['type' => 'object'])
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tools'])->toHaveCount(2)
+        ->and($decoded['tools'][0]['function']['name'])->toBe('foo')
+        ->and($decoded['tools'][1]['function']['name'])->toBe('bar');
+});
+
+test('setTools followed by setStrictTool appends without dropping existing tools', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    $preset = [[
+        'type' => 'function',
+        'function' => ['name' => 'preset_tool', 'description' => 'preset'],
+    ]];
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setTools($preset)
+        ->setStrictTool('strict_tool', ['type' => 'object'])
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect($decoded['tools'])->toHaveCount(2)
+        ->and($decoded['tools'][0]['function']['name'])->toBe('preset_tool')
+        ->and($decoded['tools'][1]['function']['name'])->toBe('strict_tool')
+        ->and($decoded['tools'][1]['function']['strict'])->toBeTrue();
+});
+
+test('setStrictTool omits the description key when not provided', function () {
+    $factory = new Psr17Factory;
+    $capturedBody = null;
+    $httpClient = v220MockClient($factory, $capturedBody);
+
+    (new DeepSeekClient($httpClient))
+        ->query('Hello')
+        ->setStrictTool('noop', ['type' => 'object'])
+        ->run();
+
+    $decoded = json_decode((string) $capturedBody, true);
+
+    expect(array_keys($decoded['tools'][0]['function']))
+        ->toBe(['name', 'parameters', 'strict']);
+});
+
+// =====================================================================
+// Enum surface
+// =====================================================================
+
+test('new QueryFlags cases exist with the expected scalar values', function () {
+    expect(QueryFlags::STOP->value)->toBe('stop')
+        ->and(QueryFlags::TOP_P->value)->toBe('top_p')
+        ->and(QueryFlags::TOOL_CHOICE->value)->toBe('tool_choice')
+        ->and(QueryFlags::LOGPROBS->value)->toBe('logprobs')
+        ->and(QueryFlags::TOP_LOGPROBS->value)->toBe('top_logprobs')
+        ->and(QueryFlags::USER->value)->toBe('user')
+        ->and(QueryFlags::THINKING->value)->toBe('thinking')
+        ->and(QueryFlags::REASONING_EFFORT->value)->toBe('reasoning_effort');
+});
+
+test('ReasoningEffort enum exposes "high" and "max"', function () {
+    expect(ReasoningEffort::HIGH->value)->toBe('high')
+        ->and(ReasoningEffort::MAX->value)->toBe('max');
+});
+
+test('ThinkingType enum exposes "enabled" and "disabled"', function () {
+    expect(ThinkingType::ENABLED->value)->toBe('enabled')
+        ->and(ThinkingType::DISABLED->value)->toBe('disabled');
+});
+
+test('ToolChoiceMode enum exposes "none", "auto", and "required"', function () {
+    expect(ToolChoiceMode::NONE->value)->toBe('none')
+        ->and(ToolChoiceMode::AUTO->value)->toBe('auto')
+        ->and(ToolChoiceMode::REQUIRED->value)->toBe('required');
+});