Showing the first 18000 characters. Load the complete file to read the remainder.
# Box Dev Docs
> Box for Developers: guides, API reference, SDKs, and tutorials for building content management, Box AI, and enterprise integrations.
- [Home (368 pages)](https://developer.box.com/_llms/en/home.txt): Documentation for Home.
- [API reference (417 pages)](https://developer.box.com/_llms/en/api-reference.txt): Documentation for API reference.
## AI
### AI
- [Build with AI](https://developer.box.com/ai.md)
- [Box Agent Skills](https://developer.box.com/ai/agent-skills.md): Find, install, and use Box Agent Skills with your AI assistant.
- [Box AI API agent configurator](https://developer.box.com/agent-configurator.md): Configure Box AI API agents with a visual configurator. Choose from multiple AI models from leading providers, all accessible through a single, easy-to-use platform. Fine-tune Box AI API and LLM parameters, then export a ready-to-use JSON configuration for your Box AI API requests.
- [Box Mount](https://developer.box.com/guides/box-mount.md): Mount a Box folder into an agent runtime as a POSIX filesystem, kept in two-way sync, with Box permissions and governance applied to every operation.
#### Supported AI models
- [Supported AI models](https://developer.box.com/guides/box-ai/ai-models/index.md): Core and customer-enabled AI models available in Box AI, with access levels, capability tiers, and API names.
- [OpenAI GPT-6.1 Sol](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-6-1-sol-model-card.md): Model card for OpenAI GPT-6.1 Sol, including API name, context window, and Box AI override details.
- [OpenAI GPT-6 Sol](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-6-sol-model-card.md): Model card for OpenAI GPT-6 Sol, including API name, context window, and Box AI override details.
- [OpenAI GPT-6 Luna](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-6-luna-model-card.md): Model card for OpenAI GPT-6 Luna, including API name, context window, and Box AI override details.
- [OpenAI GPT-6 Astra](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-6-astra-model-card.md): Model card for OpenAI GPT-6 Astra, including API name, context window, and Box AI override details.
- [OpenAI GPT-5.6 Sol](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-6-sol-model-card.md)
- [OpenAI GPT-5.6 Terra](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-6-terra-model-card.md)
- [OpenAI GPT-5.6 Luna](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-6-luna-model-card.md): Model card for OpenAI GPT-5.6 Luna, including API name, context window, and Box AI override details.
- [OpenAI GPT-5.5](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-5-model-card.md)
- [OpenAI GPT-5.4](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-4-model-card.md)
- [OpenAI GPT-5.4 mini](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-4-mini-model-card.md)
- [OpenAI GPT-5.2](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-2-model-card.md)
- [OpenAI GPT-5.1](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-1-model-card.md)
- [OpenAI GPT-5](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-model-card.md)
- [OpenAI GPT-5 mini](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-5-mini-model-card.md)
- [OpenAI GPT o3](https://developer.box.com/guides/box-ai/ai-models/openai-gpt-o3-model-card.md)
- [Azure text-embedding-ada-002](https://developer.box.com/guides/box-ai/ai-models/azure-text-embedding-ada-002-model-card.md)
- [Google Gemini 4 Argon](https://developer.box.com/guides/box-ai/ai-models/google-gemini-4-argon-model-card.md): Model card for Google Gemini 4 Argon, including API name, context window, and Box AI override details.
- [Google Gemini 3.8 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-8-flash-model-card.md)
- [Google Gemini 3.7 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-7-flash-model-card.md)
- [Google Gemini 3.6 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-6-flash-model-card.md)
- [Google Gemini 3.5 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-5-flash-model-card.md)
- [Google Gemini 3.5 Flash Lite](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-5-flash-lite-model-card.md)
- [Google Gemini 3.1 Pro](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-1-pro-model-card.md)
- [Google Gemini 3.1 Flash Lite](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-1-flash-lite-model-card.md)
- [Google Gemini 3 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-3-flash-model-card.md)
- [Google Gemini 2.5 Pro](https://developer.box.com/guides/box-ai/ai-models/google-gemini-2-5-pro-model-card.md)
- [Google Gemini 2.5 Flash](https://developer.box.com/guides/box-ai/ai-models/google-gemini-2-5-flash-model-card.md)
- [AWS Claude Haiku 5.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-haiku-5-5-model-card.md): Model card for AWS Claude Haiku 5.5, including API name, context window, and Box AI override details.
- [AWS Claude Sonnet 5.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-sonnet-5-5-model-card.md)
- [AWS Claude Opus 5.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-opus-5-5-model-card.md): Model details for AWS Claude Opus 5.5, a premium Anthropic model hosted on Amazon Web Services and available in Box AI.
- [AWS Claude Fable 5.1](https://developer.box.com/guides/box-ai/ai-models/aws-claude-fable-5-1-model-card.md)
- [AWS Claude Fable 5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-fable-5-model-card.md)
- [AWS Claude Opus 5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-opus-5-model-card.md)
- [AWS Claude Sonnet 5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-sonnet-5-model-card.md)
- [AWS Claude Opus 4.8](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-8-opus-model-card.md)
- [AWS Claude Opus 4.7](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-7-opus-model-card.md)
- [AWS Claude Opus 4.6](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-6-opus-model-card.md)
- [AWS Claude Sonnet 4.6](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-6-sonnet-model-card.md)
- [AWS Claude Opus 4.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-5-opus-model-card.md)
- [AWS Claude Sonnet 4.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-5-sonnet-model-card.md)
- [AWS Claude Haiku 4.5](https://developer.box.com/guides/box-ai/ai-models/aws-claude-4-5-haiku-model-card.md)
- [IBM Llama 4 Maverick](https://developer.box.com/guides/box-ai/ai-models/ibm-llama-4-maverick-model-card.md)
- [Meta Muse Spark 1.3](https://developer.box.com/guides/box-ai/ai-models/meta-muse-spark-1-3-model-card.md)
- [Meta Muse Spark 1.2](https://developer.box.com/guides/box-ai/ai-models/meta-muse-spark-1-2-model-card.md)
- [IBM Mistral Medium 3](https://developer.box.com/guides/box-ai/ai-models/ibm-mistral-medium-3-model-card.md)
- [IBM Mistral Small 3.1](https://developer.box.com/guides/box-ai/ai-models/ibm-mistral-small-3-1-model-card.md)
#### AI integrations
- [AI integrations](https://developer.box.com/ai/integrations.md): Connect Box to LangChain, LlamaIndex, Pinecone, Weaviate, and other AI frameworks. Use Box Hubs and Box AI Ask for permission-aware RAG.
- [Build RAG over enterprise documents stored in Box](https://developer.box.com/ai/rag.md): Build retrieval-augmented generation (RAG) over documents in Box. Start with Box Hubs and Box AI Ask for permission-aware answers. Export to Pinecone or Weaviate, or use LangChain, when you need a custom retrieval stack.
- [Box and Pinecone](https://developer.box.com/ai/vector-databases/pinecone.md): Build a RAG pipeline by connecting Box content to Pinecone for vector search and question answering.
- [Box and Weaviate](https://developer.box.com/ai/vector-databases/weaviate.md): Build an end-to-end RAG workflow by embedding Box content into Weaviate and querying it with the Weaviate Query Agent.
#### Box AI
- [Box AI](https://developer.box.com/guides/box-ai/index.md): Ask questions, extract metadata, and summarize files stored in Box with the Box AI API. Covers request handling, input limits, and agent overrides.
- [Get started with Box AI](https://developer.box.com/guides/box-ai/ai-tutorials/prerequisites.md): Create a platform app, enable the AI scope, and generate a developer token to start using the Box AI API.
- [Ask questions to Box AI](https://developer.box.com/guides/box-ai/ai-tutorials/ask-questions.md): Use the POST /ai/ask endpoint to ask questions about one or more files stored in Box.
- [Generate text with Box AI](https://developer.box.com/guides/box-ai/ai-tutorials/generate-text.md): Use the POST /ai/text_gen endpoint to generate or refine text based on file content stored in Box.
- [Extract APIs overview and use cases](https://developer.box.com/guides/box-ai/ai-tutorials/extract-use-cases.md): Explore Box AI Extract use cases including structured and freeform metadata extraction, and how to run extraction jobs across many documents in parallel with one file per request.
- [Extract metadata from file (freeform)](https://developer.box.com/guides/box-ai/ai-tutorials/extract-metadata.md): Use the POST /ai/extract endpoint to extract metadata from documents using natural language prompts.
- [Extract metadata from file (structured)](https://developer.box.com/guides/box-ai/ai-tutorials/extract-metadata-structured.md): Use the POST /ai/extract_structured endpoint to extract metadata using templates, field definitions, or the Enhanced Extract Agent.
- [Box AI API](https://developer.box.com/ai/box-ai-api.md): Ask questions, extract metadata, generate text, and summarize files in Box with the Box AI API.
##### AI model overrides
- [AI model overrides](https://developer.box.com/guides/box-ai/ai-agents/index.md): Override the default AI model, prompt, and LLM parameters used by Box AI API endpoints.
- [Get default AI agent configuration](https://developer.box.com/guides/box-ai/ai-agents/get-agent-default-config.md): Fetch the current default agent configuration for ask, text_gen, extract, and extract_structured modes.
- [AI agent configuration versioning](https://developer.box.com/guides/box-ai/ai-agents/ai-agent-versioning.md): See how Box versions AI agent snapshots, the 12-month support guarantee, and transition timelines.
- [AI model configuration override reference](https://developer.box.com/guides/box-ai/ai-agents/ai-agent-overrides.md): Full ai_agent parameter reference with sample configurations and parameter details.
- [Override AI model configuration](https://developer.box.com/guides/box-ai/ai-tutorials/default-agent-overrides.md): Step-by-step examples of overriding prompts and models for ask, text_gen, and extract endpoints.
##### AI quick starts
- [Extract structured data with Box AI in Python](https://developer.box.com/guides/box-ai/quick-start/box-ai-extract.md): Learn how to use Box AI to automatically extract structured data from documents and store it as searchable metadata using the Box Python SDK.
- [Natural language extraction with Box AI in Python](https://developer.box.com/guides/box-ai/quick-start/box-ai-extract-freeform.md): Learn how to use Box AI to automatically extract data from documents with natural language prompts using the Box Python SDK.
- [Structured extraction with the Box AI Enhanced Extract Agent in Python](https://developer.box.com/guides/box-ai/quick-start/box-ai-extract-enhanced.md): Learn how to use the Box AI Enhanced Extract Agent to automatically extract data from large, complex documents using the Box Python SDK.
- [Summarize files with Box AI in Python](https://developer.box.com/guides/box-ai/quick-start/box-ai-summarize.md): Learn how to use Box AI to automatically extract structured data from documents and store it as searchable metadata using the Box Python SDK.
##### Box AI Studio
- [Box AI Studio](https://developer.box.com/guides/ai-studio/index.md): Create, manage, and deploy custom AI agents through the Box AI Studio API.
- [Get started with AI Studio](https://developer.box.com/guides/ai-studio/getting-started-ai-studio.md): Set up a platform app and authenticate to start creating custom AI agents with the Box AI Studio API.
###### AI Studio Agents
- [AI Studio agents](https://developer.box.com/guides/ai-studio/ai-studio-agents/index.md): Create, list, update, and delete custom AI agents through the Box AI Studio API.
- [Create agents](https://developer.box.com/guides/ai-studio/ai-studio-agents/create-agents.md): Create a custom AI agent using the Box AI Studio API.
- [List agents](https://developer.box.com/guides/ai-studio/ai-studio-agents/get-agents.md): List all AI agents in your enterprise using the Box AI Studio API.
- [Get AI agent by ID](https://developer.box.com/guides/ai-studio/ai-studio-agents/get-agent-id.md): Retrieve a specific AI agent's configuration using the Box AI Studio API.
- [Update AI agents](https://developer.box.com/guides/ai-studio/ai-studio-agents/update-agents.md): Update a custom AI agent's configuration using the Box AI Studio API.
- [Delete AI agents](https://developer.box.com/guides/ai-studio/ai-studio-agents/delete-agents.md): Permanently delete a custom AI agent using the Box AI Studio API.
#### Box MCP server
- [Box MCP server](https://developer.box.com/guides/box-mcp/index.md): Hosted Model Context Protocol (MCP) server at https://mcp.box.com that lets AI agents search and use Box files over OAuth.
- [Give an AI agent secure, permission-aware access to enterprise documents](https://developer.box.com/guides/box-mcp/permission-aware-access.md): Give an AI agent secure access to enterprise documents in Box. The Box MCP server and Box APIs act as the authenticated user or application. Scopes, collaborations, and waterfall permissions still apply.
- [Available tools](https://developer.box.com/guides/box-mcp/tools.md): Reference list of Box MCP server tools by category, including which tools an admin must turn on before agents can use them.
##### Configuration
- [Set up the MCP server](https://developer.box.com/guides/box-mcp/setup.md)
###### Platform setup guides
- [Platform setup guides](https://developer.box.com/guides/box-mcp/integrations/index.md): List of AI platforms with step-by-step setup guides for the hosted Box MCP server.
- [Anthropic's Messages API](https://developer.box.com/guides/box-mcp/integrations/anthropic-messages-api.md)
- [Box MCP for Claude Code](https://developer.box.com/guides/box-mcp/integrations/claude-code.md): Connect the Box MCP server to Claude Code so your coding agent can search and use Box content from the terminal.
- [Codex](https://developer.box.com/guides/box-mcp/integrations/codex.md)
- [Cursor](https://developer.box.com/guides/box-mcp/integrations/cursor.md)
- [Box MCP for GitHub Copilot](https://developer.box.com/guides/box-mcp/integrations/github-copilot.md): Connect the Box MCP server to GitHub Copilot in VS Code so Copilot can search and use files in Box.
- [Groq Responses API](https://developer.box.com/guides/box-mcp/integrations/groq-responses-api.md)
- [LangSmith Agent Builder](https://developer.box.com/guides/box-mcp/integrations/langsmith-agent-builder.md)
- [Microsoft Foundry Tools](https://developer.box.com/guides/box-mcp/integrations/microsoft-foundry-tools.md): Connect the Box MCP server to Microsoft Foundry Tools. The agent acts as the authorizing Box user. Scopes and Box file permissions still apply.
- [OpenAI AgentKit](https://developer.box.com/guides/box-mcp/integrations/openai-agentkit.md)
- [Windows ODR (On Device Registry)](https://developer.box.com/guides/box-mcp/integrations/windows-on-device-registry.md)
##### Appendix
- [Self-hosted Box MCP server (legacy)](https://developer.box.com/guides/box-mcp/self-hosted/index.md): Deprecated open-source MCP server for local use. Prefer the remote Box MCP server for new integrations.
## Tools
### Dev tools
- [Tools](https://developer.box.com/sdks-and-tools.md): The following SDKs and tools are developed and supported by Box
- [The `box` npm package](https://developer.box.com/guides/tooling/box-npm-package.md): What `npm install box` provides, how to use the bundled SDK and CLI, and what existing `box-node-sdk` users need to know.
- [Box Mount](https://developer.box.com/guides/box-mount.md): Mount a Box folder into an agent runtime as a POSIX filesystem, kept in two-way sync, with Box permissions and governance applied to every operation.
- [Box Embed](https://developer.box.com/guides/embed/box-embed.md)
#### SDKs
- [Box SDKs](https://developer.box.com/guides/tooling/sdks/index.md)
- [Box SDK versioning strategy](https://developer.box.com/guides/tooling/sdks/sdk-versioning.md)
- [Install .NET SDK](https://developer.box.com/guides/tooling/sdks/dotnet.md)
- [Box Java SDK](https://developer.box.com/guides/tooling/sdks/java.md): Install the Box Java SDK with Maven or Gradle and start calling the Box API from a Java application.
- [Install Python SDK](https://developer.box.com/guides/tooling/sdks/python.md)
- [Install Node SDK](https://developer.box.com/guides/tooling/sdks/node.md)
- [Install Salesforce SDK](https://developer.box.com/guides/tooling/sdks/salesforce.md)
##### Deprecated
- [Deprecated Box Next Gen TypeScript SDK](https://developer.box.com/guides/tooling/sdks/typescript-gen.md)
- [Deprecated Box Next Gen Python SDK](https://developer.box.com/guides/tooling/sdks/python-gen.md)
- [Deprecated Box Next Gen .NET SDK](https://developer.box.com/guides/tooling/sdks/dotnet-gen.md)
- [Deprecated Box Next Gen Java SDK](https://developer.box.com/guides/tooling/sdks/java-gen.md)
- [Deprecated Box Next Gen Swift SDK](https://developer.box.com/guides/tooling/sdks/swift-gen.md)
#### CLI
- [Box CLI](https://developer.box.com/gu
Measured from the saved file
Linked document
UTF-8 text size
29.7 KB30405 bytes
Token count
7848o200k_base encoding
Markdown links
210207 distinct URLs
Fenced code blocks
0316 lines in the file
Size measures the captured UTF-8 text, not compressed transfer size. Tokens are counted with a named tokenizer; other models can count differently. Neither measure establishes context compatibility or answer quality.
Project H1
Box Dev Docs
Opening blockquote
Box for Developers: guides, API reference, SDKs, and tutorials for building content management, Box AI, and enterprise integrations.
Context architecture
How the content is divided
7 H2 sections organize 210 parsed Markdown links. 195 links point to .md or .markdown paths; 5 link occurrences use another hostname. Linked destinations have not been recursively validated.
H2 section
Links
UTF-8 bytes
Tokens
AILine 8
96
15754
4308
ToolsLine 142
75
8106
2034
TutorialsLine 264
15
3921
797
ChangelogLine 282
9
1082
307
OpenAPI SpecsLine 294
3
139
57
OptionalLine 300 · Optional
5
348
92
IndexesLine 310
5
612
152
Link counts exclude parsed fenced and inline code and resolve relative destinations against the final endpoint. Section sizes cover the text below each H2 until the next H2.
Embedded material
Code, installation and API examples
Fenced blocks
0 closed
Declared languages
No fenced blocks detected
Package-install commands
1 detected in fenced or inline code
Authentication examples
0 fenced lines match API-key, Authorization, Bearer, OAuth or access-token patterns
HTTP request examples
0 fenced lines match request-method patterns
Authentication links
1 links have authentication-related labels
Embedded example text and links to further documentation are counted separately. A code block does not establish that a sample runs or that an authentication flow succeeds.
Detected installation lines
Source line 147 · inline
npm install box
Publisher prioritization
The Optional section
1 H2 section named “Optional” contain 5 parsed links. 205 links sit outside those sections. That puts 2.4% of the file’s parsed links under Optional.
The community proposal uses Optional by convention for secondary links that can be skipped for a shorter context. This label does not establish that everything outside it is critical, or that the material inside it is deprecated.
Machine-readable routes
MCP references and configuration text
MCP-related text was detected in the captured file. 16 links have MCP-related labels or paths. 0 fenced lines match mcpServers, mcp_servers or an mcp add command.
Whole-file size not established by the bounded response
Token count
Not measured for the whole file
Retrieval scope
Bounded prefix inspected; the whole body was not downloaded
A verified text response establishes a companion at the checked URL. It does not prove the bundle contains all product documentation, that generation is automated, or that it fits an assistant’s context window.
Change evidence
What can be said about freshness
This capture establishes a baseline. Update frequency cannot be determined from one observation.
Linked release and migration material
14 links have release, changelog, deprecation or migration terms in their labels or paths. These provide places to investigate product changes separately.
File changes can reveal documentation edits or navigation changes. They do not by themselves measure product release velocity, accuracy, agent adoption or AI visibility.
Counts use a deterministic parser, not a spec-compliance grade. Unusual Markdown and indented code can fall outside its recognized patterns. See the llms.txt proposal and tokenizer implementation.