docs: align LLM model docs and pricing

docs/llmservice/models/gemini-3-5-flash-lite.md

@@ -0,0 +1,48 @@
+# Gemini 3.5 Flash-Lite
+
+## Overview
+
+Gemini 3.5 Flash-Lite is a Gemini 3-series model for low-latency, high-throughput multimodal processing. It is suited to document extraction, classification, translation, structured data tasks, and parallel subagent workflows. On B.AI, use the model ID `gemini-3.5-flash-lite`.
+
+## Key Features
+
+* **High-Throughput Processing:** Designed for latency-sensitive, high-volume production traffic.
+* **Multimodal Input:** Supports text, image, video, audio, and PDF input with text output.
+* **Configurable Thinking:** Supports multiple thinking levels, with a lightweight default for fast responses.
+* **Long Context:** Supports up to 1,048,576 input tokens for large-scale document and data processing.
+* **Tool Integration:** Supports function calling, structured outputs, code execution, grounding, URL context, and file-search workflows when enabled by the API configuration.
+
+## Best Use Cases
+
+* Document and data extraction from PDFs, records, tables, and other multimodal inputs.
+* Classification, routing, translation, tagging, and summarization at scale.
+* Parallel subagent tasks that benefit from fast responses and structured outputs.
+* Interactive applications, including chatbots and high-throughput content workflows.
+
+## Capabilities and Limitations
+
+| Capability | Detailed Description |
+| :--- | :--- |
+| **Reasoning** | Supports configurable thinking levels for trading off response speed and reasoning depth. |
+| **Coding** | Suitable for focused coding assistance, tool-using tasks, and structured-output workflows. |
+| **Multimodal** | Supports text, image, video, audio, and PDF input with text output. |
+| **Context Window** | Supports up to 1,048,576 input tokens. |
+| **Max Output** | Supports up to 65,536 output tokens. |
+| **Tool Use** | Supports function calling, structured outputs, code execution, grounding, URL context, and file-search workflows when available. |
+| **Multilingual** | Suitable for translation and multilingual production workloads. |
+
+### Known Limitations
+
+* The model can produce inaccurate responses; validate extracted or classified information before using it in critical workflows.
+* Actual latency and tool availability vary with thinking level, prompt size, media inputs, and API configuration.
+* Output is text-only; native image and audio generation are not included.
+
+## Credits Usage
+
+| Model | Input (Credits/Token) | Cache Write (Credits/Token) | Cache Read (Credits/Token) | Output (Credits/Token) | Web Search (Credits/Use) | Billing Notes |
+| :--- | --------------------: | --------------------------: | -------------------------: | ---------------------: | -----------------------: | :--- |
+| **Gemini 3.5 Flash-Lite** | `0.30` | `0.30` | `0.03` | `2.50` | `14,000` | - |
+
+:::info Pricing note
+Prices shown in the documentation are B.AI standard reference prices for base billing purposes. B.AI may provide lower actual usage costs through top-up bonuses and account benefits. Specific prices, bonus Credits, and account benefits are subject to the platform display and final billing records.
+:::

docs/llmservice/models/gemini-3-6-flash.md

@@ -0,0 +1,48 @@
+# Gemini 3.6 Flash
+
+## Overview
+
+Gemini 3.6 Flash is a Gemini 3-series model for agentic coding, multimodal knowledge work, spatial reasoning, and multi-step workflows. On B.AI, use the model ID `gemini-3.6-flash`.
+
+## Key Features
+
+* **Agentic Workflows:** Designed for repeated reasoning, tool calls, and multi-step task execution.
+* **Multimodal Input:** Supports text, image, video, audio, and PDF input with text output.
+* **Configurable Thinking:** Supports thinking levels for balancing quality, latency, and token usage.
+* **Long Context:** Supports up to 1,048,576 input tokens for large document collections, codebases, and extended task histories.
+* **Tool Integration:** Supports function calling, structured outputs, code execution, grounding, URL context, and file-search workflows when enabled by the API configuration.
+
+## Best Use Cases
+
+* Agentic software engineering, including code generation, debugging, migrations, and repository-scale changes.
+* Multimodal knowledge work, such as document analysis, chart interpretation, and report drafting.
+* Long-context analysis across large codebases, document sets, and task histories.
+* Tool-using workflows that combine reasoning, structured output, and repeated execution steps.
+
+## Capabilities and Limitations
+
+| Capability | Detailed Description |
+| :--- | :--- |
+| **Reasoning** | Supports configurable thinking levels for tasks that need different quality, latency, and cost tradeoffs. |
+| **Coding** | Suitable for iterative coding, debugging, and multi-step engineering workflows. |
+| **Multimodal** | Supports text, image, video, audio, and PDF input with text output. |
+| **Context Window** | Supports up to 1,048,576 input tokens. |
+| **Max Output** | Supports up to 65,536 output tokens. |
+| **Tool Use** | Supports function calling, structured outputs, code execution, grounding, URL context, and file-search workflows when available. |
+| **Multilingual** | Suitable for multilingual developer and knowledge-work tasks. |
+
+### Known Limitations
+
+* The model can produce inaccurate responses; use grounding or current external data when up-to-date information is required.
+* Actual latency and tool availability vary with thinking level, prompt size, media inputs, and API configuration.
+* Output is text-only; native image and audio generation are not included.
+
+## Credits Usage
+
+| Model | Input (Credits/Token) | Cache Write (Credits/Token) | Cache Read (Credits/Token) | Output (Credits/Token) | Web Search (Credits/Use) | Billing Notes |
+| :--- | --------------------: | --------------------------: | -------------------------: | ---------------------: | -----------------------: | :--- |
+| **Gemini 3.6 Flash** | `1.50` | `1.50` | `0.15` | `7.50` | `14,000` | - |
+
+:::info Pricing note
+Prices shown in the documentation are B.AI standard reference prices for base billing purposes. B.AI may provide lower actual usage costs through top-up bonuses and account benefits. Specific prices, bonus Credits, and account benefits are subject to the platform display and final billing records.
+:::

docs/llmservice/models/grok-4.5.md

@@ -0,0 +1,48 @@
+# Grok 4.5
+
+## Overview
+
+Grok 4.5 is a SpaceXAI reasoning model for coding, agentic tasks, engineering, and knowledge work. On B.AI, use the model ID `grok-4.5`.
+
+## Key Features
+
+* **Agentic Engineering:** Suitable for multi-step coding, debugging, refactoring, and repository-scale work.
+* **Configurable Reasoning:** Supports reasoning-effort controls for balancing response speed and depth.
+* **Long Multimodal Context:** Supports text and image input within a 500,000-token context window and returns text output.
+* **Tool-Using Workflows:** Supports function calling and structured outputs, with additional tool capabilities depending on API configuration.
+* **Long-Context Pricing:** Requests with more than 200,000 prompt tokens use the long-context pricing tier.
+
+## Best Use Cases
+
+* Repository-scale coding, including implementation, debugging, refactoring, and technical review.
+* Tool-using agents that combine structured output, function calling, and repeated reasoning loops.
+* Technical analysis across engineering, science, mathematics, and mixed text/image materials.
+* Long-context document and knowledge-work workflows.
+
+## Capabilities and Limitations
+
+| Capability | Detailed Description |
+| :--- | :--- |
+| **Reasoning** | Supports configurable reasoning effort for tasks with different latency and depth requirements. |
+| **Coding** | Suitable for multi-step software engineering and technical problem-solving workflows. |
+| **Multimodal** | Supports text and image input with text output. |
+| **Context Window** | Supports up to 500,000 tokens. |
+| **Tool Use** | Supports function calling and structured outputs; tool availability depends on API configuration. |
+| **Multilingual** | Suitable for natural-language technical and knowledge-work requests. |
+
+### Known Limitations
+
+* The model can produce inaccurate responses; validate results before using them in critical workflows.
+* Requests with more than 200,000 prompt tokens use the long-context pricing tier for the request.
+* Actual tool availability and related charges depend on the real-time options exposed by the platform.
+
+## Credits Usage
+
+| Model and Context Tier | Input (Credits/Token) | Cache Write (Credits/Token) | Cache Read (Credits/Token) | Output (Credits/Token) |
+| :--- | --------------------: | --------------------------: | -------------------------: | ---------------------: |
+| **Grok 4.5** | `2.00` | `2.00` | `0.30` | `6.00` |
+| **Grok 4.5** (>200K prompt tokens) | `4.00` | `4.00` | `0.60` | `12.00` |
+
+:::info Pricing note
+Prices shown in the documentation are B.AI standard reference prices for base billing purposes. B.AI may provide lower actual usage costs through top-up bonuses and account benefits. Specific prices, bonus Credits, and account benefits are subject to the platform display and final billing records.
+:::

docs/llmservice/pricing-and-usage.md

@@ -25,7 +25,7 @@ The platform uses a unified Credits system to measure and settle usage across al
 | DeepSeek V3.2     |                  0.29 |                        0.29 |                      0.145 |                   0.44 |                        - |
 | DeepSeek V4 Flash |                  0.28 |                        0.28 |                     0.0056 |                   0.56 |                        - |
 | DeepSeek V4 Pro   |                  0.87 |                        0.87 |                     0.0087 |                   1.74 |                        - |
-| Grok-4.5          |                  2.00 |                        2.00 |                       0.30 |                   6.00 |                        - |
+| Grok 4.5          |                  2.00 |                        2.00 |                       0.30 |                   6.00 |                        - |
 | GPT-5.6 Sol       |                  5.00 |                        6.25 |                       0.50 |                  30.00 |                   10,000 |
 | GPT-5.6 Terra     |                  2.50 |                       3.125 |                       0.25 |                  15.00 |                   10,000 |
 | GPT-5.6 Luna      |                  1.00 |                        1.25 |                       0.10 |                   6.00 |                   10,000 |
@@ -47,8 +47,10 @@ The platform uses a unified Credits system to measure and settle usage across al
 | Claude Sonnet 4.6 |                  3.00 |                        3.75 |                       0.30 |                  15.00 |                   10,000 |
 | Claude Sonnet 4.5 |                  3.00 |                        3.75 |                       0.30 |                  15.00 |                   10,000 |
 | Claude Haiku 4.5  |                  1.00 |                        1.25 |                       0.10 |                   5.00 |                   10,000 |
-| Gemini 3.1 Pro    |                  2.00 |                        2.00 |                       0.20 |                  12.00 |                   14,000 |
+| Gemini 3.6 Flash  |                  1.50 |                        1.50 |                       0.15 |                   7.50 |                   14,000 |
 | Gemini 3.5 Flash  |                  1.50 |                        1.50 |                       0.15 |                   9.00 |                   14,000 |
+| Gemini 3.5 Flash-Lite |               0.30 |                        0.30 |                       0.03 |                   2.50 |                   14,000 |
+| Gemini 3.1 Pro    |                  2.00 |                        2.00 |                       0.20 |                  12.00 |                   14,000 |
 | Gemini 3 Flash    |                  0.50 |                        0.50 |                       0.05 |                   3.00 |                   14,000 |
 
 :::caution Main table scope

i18n/zh-Hans/docusaurus-plugin-content-docs/current/llmservice/models/gemini-3-5-flash-lite.md

@@ -0,0 +1,48 @@
+# Gemini 3.5 Flash-Lite
+
+## 概述
+
+Gemini 3.5 Flash-Lite 是 Gemini 3 系列模型，面向低延迟、高吞吐的多模态处理任务。它适用于文档提取、分类、翻译、结构化数据处理和并行子智能体工作流。在 B.AI 中使用模型 ID：`gemini-3.5-flash-lite`。
+
+## 核心特性
+
+* **高吞吐处理：** 面向延迟敏感、高调用量的生产工作负载。
+* **多模态输入：** 支持文本、图片、视频、音频和 PDF 输入，输出为文本。
+* **可配置思考：** 支持多个思考等级，并提供轻量级默认配置以获得更快响应。
+* **长上下文：** 最多支持 1,048,576 个输入 token，适合大规模文档和数据处理。
+* **工具集成：** 在 API 配置启用时，支持函数调用、结构化输出、代码执行、联网检索、URL 上下文和文件搜索工作流。
+
+## 推荐场景
+
+* 从 PDF、记录、表格和其他多模态内容中提取文档和数据。
+* 大规模分类、路由、翻译、标签和摘要任务。
+* 需要快速响应和结构化输出的并行子智能体任务。
+* 聊天机器人和高吞吐内容工作流等交互式应用。
+
+## 能力与限制
+
+| 能力 | 说明 |
+| :--- | :--- |
+| **推理** | 支持可配置的思考等级，用于平衡响应速度和推理深度。 |
+| **编程** | 适用于聚焦型编程辅助、工具调用任务和结构化输出工作流。 |
+| **多模态** | 支持文本、图片、视频、音频和 PDF 输入，输出为文本。 |
+| **上下文窗口** | 最多支持 1,048,576 个输入 token。 |
+| **最大输出** | 最多支持 65,536 个输出 token。 |
+| **工具调用** | 在可用时支持函数调用、结构化输出、代码执行、联网检索、URL 上下文和文件搜索工作流。 |
+| **多语言** | 适用于翻译和多语言生产工作负载。 |
+
+### 已知限制
+
+* 模型可能生成不准确内容；关键工作流中的提取或分类结果应先进行校验。
+* 实际延迟和工具可用性会受思考等级、提示词长度、媒体输入和 API 配置影响。
+* 输出仅支持文本，不包含原生图片或音频生成能力。
+
+## Credits 用量
+
+| 模型 | 输入（Credits/Token） | 缓存写入（Credits/Token） | 缓存读取（Credits/Token） | 输出（Credits/Token） | 网页搜索（Credits/次） | 计费说明 |
+| :--- | --------------------: | -------------------------: | -------------------------: | --------------------: | ----------------------: | :--- |
+| **Gemini 3.5 Flash-Lite** | `0.30` | `0.30` | `0.03` | `2.50` | `14,000` | - |
+
+:::info 价格说明
+文档价格为 B.AI 平台模型标准参考价，仅供基础计费说明使用。B.AI 可能会通过充值赠送及账户权益等方式，为用户提供更低的实际使用成本。具体价格、赠送积分及账户权益请以平台页面展示及最终账单为准。
+:::

i18n/zh-Hans/docusaurus-plugin-content-docs/current/llmservice/models/gemini-3-6-flash.md

@@ -0,0 +1,48 @@
+# Gemini 3.6 Flash
+
+## 概述
+
+Gemini 3.6 Flash 是 Gemini 3 系列模型，适用于智能体编程、多模态知识工作、空间推理和多步骤工作流。在 B.AI 中使用模型 ID：`gemini-3.6-flash`。
+
+## 核心特性
+
+* **智能体工作流：** 适合重复推理、工具调用和多步骤任务执行。
+* **多模态输入：** 支持文本、图片、视频、音频和 PDF 输入，输出为文本。
+* **可配置思考：** 支持思考等级配置，用于平衡质量、延迟和 token 消耗。
+* **长上下文：** 最多支持 1,048,576 个输入 token，适用于大型文档集、代码库和长任务历史。
+* **工具集成：** 在 API 配置启用时，支持函数调用、结构化输出、代码执行、联网检索、URL 上下文和文件搜索工作流。
+
+## 推荐场景
+
+* 智能体软件工程，包括代码生成、调试、迁移和仓库级修改。
+* 多模态知识工作，例如文档分析、图表解读和报告撰写。
+* 面向大型代码库、文档集和任务历史的长上下文分析。
+* 结合推理、结构化输出和重复执行步骤的工具调用工作流。
+
+## 能力与限制
+
+| 能力 | 说明 |
+| :--- | :--- |
+| **推理** | 支持可配置的思考等级，适配不同质量、延迟和成本要求。 |
+| **编程** | 适用于迭代式编程、调试和多步骤工程工作流。 |
+| **多模态** | 支持文本、图片、视频、音频和 PDF 输入，输出为文本。 |
+| **上下文窗口** | 最多支持 1,048,576 个输入 token。 |
+| **最大输出** | 最多支持 65,536 个输出 token。 |
+| **工具调用** | 在可用时支持函数调用、结构化输出、代码执行、联网检索、URL 上下文和文件搜索工作流。 |
+| **多语言** | 适用于多语言开发和知识工作任务。 |
+
+### 已知限制
+
+* 模型可能生成不准确内容；需要最新信息时，请使用联网检索或其他可靠外部数据源。
+* 实际延迟和工具可用性会受思考等级、提示词长度、媒体输入和 API 配置影响。
+* 输出仅支持文本，不包含原生图片或音频生成能力。
+
+## Credits 用量
+
+| 模型 | 输入（Credits/Token） | 缓存写入（Credits/Token） | 缓存读取（Credits/Token） | 输出（Credits/Token） | 网页搜索（Credits/次） | 计费说明 |
+| :--- | --------------------: | -------------------------: | -------------------------: | --------------------: | ----------------------: | :--- |
+| **Gemini 3.6 Flash** | `1.50` | `1.50` | `0.15` | `7.50` | `14,000` | - |
+
+:::info 价格说明
+文档价格为 B.AI 平台模型标准参考价，仅供基础计费说明使用。B.AI 可能会通过充值赠送及账户权益等方式，为用户提供更低的实际使用成本。具体价格、赠送积分及账户权益请以平台页面展示及最终账单为准。
+:::