Services / Anthropic / Claude Platform
Claude Platform

Build Claude into your products, and run agents in production.

The Claude Platform is the second way Anthropic describes to bring Claude into an organisation: give developers the Claude API to build AI into products, services and agents. Anthropic frames it in three layers: primitives such as the Messages API and its tools; harnesses and infrastructure, including Claude Managed Agents, to ship agents that run for hours; and the controls to deploy and govern agents across the organisation.

Explore the platform →Which model?

Available directly from Anthropic, or through Amazon Web Services, Google Cloud and Microsoft Foundry with your existing cloud account and billing.

Illustration of the Claude Console, simplified. Sample data.
What it does

Models, harnesses, context and infrastructure that work together.

Anthropic's pitch for the platform is that the pieces are designed together: the models, the tools they use, the way context is managed and the infrastructure that runs agents.

01A model for every job
Fable for multi-day autonomous work, Opus as the daily driver for coding and enterprise agents, Sonnet for well-scoped work at scale, Haiku for real-time and high-volume tasks.
02Tools Claude can use
Tool use, computer use, code execution, web search and fetch, citations and files, with structured outputs that conform to your JSON schema.
03Agents without the plumbing
Claude Managed Agents runs long-running agents in managed infrastructure, with memory, scheduling, vaults for credentials and session tracing.
04Your systems and knowledge
The Model Context Protocol connects Claude to your tools and data, and skills package your procedures so Claude follows them.
05Cost under control
Batch processing, prompt caching, an effort setting and an advisor strategy trade cost and speed against intelligence, task by task.
06Enterprise data controls
No training on commercial customer data by default, regional data residency on your cloud provider, and zero data retention for eligible customers.
Explore

Six parts of the platform, up close.

Pick one to see what Anthropic offers, and how it might look in use.

The Messages API and the tools Claude can use

Full control over every layer.

  • The Messages API gives access to Claude's models with full control and customisation.
  • Tool use lets Claude call external tools and APIs; computer use lets it see and operate a browser or desktop for applications that have no API.
  • Code execution runs Python, creates visualisations and analyses data inside API calls.
  • Structured outputs make responses conform to your JSON schema.
  • Citations ground responses in source documents, and the Files API lets you upload and reference documents across conversations.
  • Web search and fetch bring current information from the web into Claude.
What you’re looking at

Computer use, for the systems with no API

Older internal applications often have no API. Computer use lets an agent work their screens, which is useful exactly where integration is hardest.

Illustration of a screen, simplified. Sample data.
Which model?

Four models, four kinds of job.

How Anthropic positions each model on the Claude Platform. Prices change; Anthropic's pricing is linked below.

ModelWhat Anthropic says it is for
Fable 5.1Next-generation intelligence for long-running agents: multi-day autonomous projects and expert-level work for frontier research.
Opus 5.5The daily driver for agentic coding and enterprise work: long-horizon coding, large migrations and multi-step agents across enterprise tools and data.
Sonnet 5.5A high-performance model for coding and agents: everyday coding and customer-facing agents and multi-tool workflows at scale.
Haiku 5.5The fastest, most cost-effective model: real-time, latency-sensitive experiences and sub-agents inside larger systems.

Claude Platform models and pricing

Managed Agents templates

Where agents start.

Anthropic's starting templates for Claude Managed Agents, each with tools, connections and permissions already set. They show the shape of work agents are being used for.

01Deep researcher
Multi-step web research with source synthesis and citations.
02Contract tracker
Extracts clauses, sets deadline reminders and tracks obligations.
03Support agent
Answers customer questions from your docs and knowledge base, and escalates when needed.
04Data analyst
Loads, explores and visualises data, builds reports and answers questions from datasets.
05Incident commander
Triages an alert, opens an incident ticket and runs the incident room.
06Structured extractor
Parses unstructured text into a typed JSON schema.

Anthropic also lists a field monitor, a sprint retro facilitator and a support-to-engineering escalator. Its named use cases include code migration, security remediation, experimental research and compliance audits.

Built for enterprise workloads

Your data, and where it is processed.

What the Claude Platform page and the cloud pages say about data.

ControlWhat Anthropic says
TrainingBy default, Anthropic does not use customer data from commercial deployments to train Claude.
ResidencyOn your cloud provider, choose where data is stored and where requests are processed, with regions in Asia-Pacific, Canada, Europe and the United States.
Zero data retentionEnterprise Frontier Safeguards give eligible customers zero data retention, with safeguards for detecting misuse.
CredentialsIn Managed Agents, vaults add secrets at the network boundary, so the agent only ever handles a placeholder.
Who uses it

The people who build with Claude.

Roles drawn from Anthropic's Platform, Managed Agents and Console pages.

RoleWhat they do
Product engineersBuild AI features into applications with the Messages API, tools and structured outputs.
Platform teamsRun agents in production with Managed Agents, vaults, sandboxes and session tracing.
ArchitectsChoose models per task, design multi-agent systems and decide which cloud Claude runs on.
FinOps and engineering managersWatch usage, cost and caching by model and key, and set limits per workspace.
Security teamsReview data residency, retention and credential handling before launch.

© 2026 Mawkish Technologies. This guide, its diagrams and illustrations are protected by copyright. You’re welcome to learn from it and to quote it with credit and a link. Reproducing it, by hand or through AI, and publishing it as your own is not allowed. Read our copyright notice

Questions

Straight answers.

What is the Claude Platform?

Anthropic's developer platform: the Claude API, its tools and Claude Managed Agents, used to build AI into products, services and agents. It is available directly and through AWS, Google Cloud and Microsoft Foundry.

What is Claude Managed Agents?

Anthropic describes it as a suite of composable APIs for building and deploying cloud-hosted agents at scale, pairing an Anthropic-managed harness with production infrastructure for state, memory, permissions and scheduled runs. It is in beta.

Can Claude run inside our own cloud account?

Yes. Anthropic offers Claude through Amazon Bedrock and the Claude Platform on AWS, on Google Cloud (managed by Anthropic or by Google), and in Microsoft Foundry, each billed through that cloud and counting towards your commitments.

Can we keep data in a particular region?

Anthropic's regional compliance page says regional data residency is available in Europe, the United States, Canada and Asia-Pacific, including Japan, South Korea, Singapore, India and Australia, through regional endpoints on AWS Bedrock, Google Cloud and Microsoft Foundry.

Is our API data used for training?

Anthropic says that by default it does not use customer data from commercial deployments to train Claude.

Which model should we start with?

It depends on the job. Anthropic positions Opus 5.5 as the daily driver for coding and enterprise agents, Sonnet 5.5 for well-scoped work at scale and Haiku 5.5 for real-time and high-volume tasks. Our models page compares them.

Checked against Anthropic's own pages in November 2026.

Across Claude

You are on Claude Platform. Where next?

Claude Enterprise, the Claude Platform, Claude and Claude Code each have their own page. Visit them in any order. Choose one to go there.

You are currently viewing Claude Platform

Also on Claude

How it all fits together

Talk to us

Talk it through with us.

Tell us what you want Claude to do for your organisation, and we will help you work out where to start, safely.

  • Your goals, data and the systems involved
  • Claude Enterprise, the Platform, or both
  • A sensible first step

Anthropic, Claude, Claude Code and the other product names on these pages are trademarks of Anthropic, PBC. No Anthropic logo or artwork is used, and nothing here implies Anthropic's sponsorship or endorsement. Other names are the property of their owners. Screens shown are our own illustrations with sample data, not Anthropic screenshots. Product descriptions follow Anthropic's own published material and may change; check claude.com and anthropic.com for current details.

Our Tevbrook Brief is the name we give to how we handle Claude Platform: one accountable partner, from the first meeting to long after go-live.

© 2026 Mawkish Technologies. This guide, its diagrams and illustrations are protected by copyright. You’re welcome to learn from it and to quote it with credit and a link. Reproducing it, by hand or through AI, and publishing it as your own is not allowed. Read our copyright notice