Small print.
Wide open.

Table of Contents

The Claude Platform

Customers choose Claude for more than the models. The Claude Platform lets developers and businesses build, 
manage, and deploy agentic AI applications on infrastructure with high reliability, security, and availability.

Platform Features

Platform features enable developers to customize their usage of Claude, including:

•

•

•

•

•

•

•

Extended  Thinking.  Claude  can  perform  additional  internal  reasoning  before  producing  a  final 
response,  working  through  complex,  multi-step  problems  to  improve  response  quality  on  tasks  like 
analysis, math, and coding.

Adaptive  Thinking.  Claude  automatically  calibrates  whether  a  given  request  calls  for  extended 
reasoning  and  how  much  reasoning  to  apply—so  developers  get  the  benefits  of  extended  reasoning 
without configuring reasoning levels request by request.

Effort Controls.  Developers  can  specify  an  effort  level  on  each  request  to  guide  how  much  reasoning 
Claude applies, allowing them to balance response quality against cost and latency.

Tool Use. Claude can natively call external functions and APIs throughout a conversation, with support 
for parallel execution and interleaved thinking between tool calls. The search tool loads only the tools 
required for a given task on demand, which can reduce token overhead compared to loading full tool 
libraries upfront.

Long  Context.  Claude  supports  large  context  windows  and  maintains  its  strong  performance  even  as 
inputs grow, meaning that developers can include entire codebases, large document sets, or long agent 
histories in a single request.

Structured Outputs. Developers can constrain Claude’s responses to a specified schema, such as valid 
JSON  (JavaScript  Object  Notation),  so  results  integrate  reliably  with  downstream  systems  and 
application code with less additional parsing or validation.

Code Execution. A sandboxed execution environment where Claude can write, run, and iterate on code 
within  an  agent  loop,  which  is  foundational  for  data  analysis  and  automation  workflows.  This 
environment  lets  Claude  call  tools  and  process  large  volumes  of  data  more  efficiently  and  run  the 
scripts packaged in Agent Skills.

• Web  Search.  A  native  tool  giving  Claude  access  to  real-time  web  information,  enabling  grounded, 

citation-aware responses in production applications.

•

Files API. Allows developers to upload documents once and reference them across multiple sessions, 
simplifying workflows that depend on persistent documents.

Context Management

Developers can connect context to Claude for their agents and applications using context primitives.

• Model Context Protocol. MCP is an open standard we introduced in November 2024 and donated to 
the Linux Foundation’s Agentic AI Foundation in December 2025. MCP defines a universal interface 
through  which  Claude,  as  well  as  other  AI  tools,  can  connect  to  enterprise  data  sources,  tools,  and 
workflows. Claude Platform’s native MCP connector handles connection management, tool discovery, 
and error handling automatically. 

•

Agent  Skills.  A  modular  framework  that  lets  organizations  encode  repeatable  tasks,  workflows,  and 
domain expertise into reusable packages, or skills, that Claude loads on demand. We provide pre-built 
skills  for  common  formats  like  Word,  Excel,  PowerPoint,  and  PDF,  while  custom  skills  can  capture 
anything from engineering standards and regulatory processes to proprietary toolchains. These skills 
turn institutional knowledge into agent capabilities.

200