# Anthropic

> Source: https://agentsindex.ai/anthropic

AI model provider for Claude and managed-agent infrastructure.

- Best for: Teams that need frontier AI for coding, agents, and enterprise workflows with safety controls.

Starting price

$0.10 per 1m input tokens

From its own pages, checked 30 October 2026

- MCP server Yes From github.com/anthropics/claude-agent-sdk-python, 29 October 2026 [github.com](https://github.com/anthropics/claude-agent-sdk-python)
- API Not stated
- Open source Yes From github.com/anthropics/claude-agent-sdk-python, 29 October 2026 [github.com](https://github.com/anthropics/claude-agent-sdk-python)
- Self-host Not stated

[Pricing](https://www.anthropic.com/pricing)

What we checked

- Pages last read: 4 November 2026
- Facts checked: 2 facts · From 2 sources, read 29 October 2026
- AI answers: Too few read · 58 buyer-question answers read, November 2026

[How we check](https://agentsindex.ai/methodology) [Editorial policy](https://agentsindex.ai/editorial-policy#not-for-sale)

[Visit Anthropic](https://agentsindex.ai/go/anthropic) [Compare](https://agentsindex.ai/compare?cat=ai-model-providers&products=anthropic)

## At a glance

- Category: [AI model providers](https://agentsindex.ai/categories/ai-model-providers)

### Company

- Founded: 2021

### For developers

- SDKs: Python, TypeScript, C#, Go, Java, PHP, Ruby
- Runs: Cloud
- Works with: Slack, Microsoft 365, Google Cloud’s Vertex AI, Microsoft Foundry, Claude on AWS, Google Cloud and 4 more

### Security

- Compliance: SOC 2 Type II, ISO 27001, ISO 42001

## What is Anthropic?

Anthropic is an AI platform for engineering, product, operations, security, and research teams that turns frontier models into systems for coding, agents, and enterprise work. It pairs Claude Opus 4.7 with Claude Code, AI agents, Claude Security, and Connectors, and is available through Claude.ai, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. Named adopters include Google Cloud, AWS, Ramp, Replit, and Cursor.

## Claude covers direct prompting and managed agents

Anthropic's developer platform has two primary paths. The [Messages API](https://platform.claude.com/docs/en/intro) gives direct model-prompting access for custom agent loops and fine-grained control. Claude Managed Agents is a configurable agent harness that runs in managed infrastructure for long-running and asynchronous work. The same introduction says current Claude models accept text and image input, return text, and support multilingual work, vision, and tool use. This is a hosted service rather than model weights delivered for local deployment. The API overview also distinguishes the first-party API from access through AWS, Google Cloud, and Microsoft Azure. Feature availability can vary across those routes, so an implementation that needs a particular capability has to check the platform-specific documentation rather than assume parity.

## The API has a conventional documented web interface

The [API overview](https://platform.claude.com/docs/en/api/overview) identifies the Claude API as a RESTful API at api.anthropic.com and says it provides programmatic access to models and Claude Managed Agents. Its listed surfaces include Messages, Message Batches, Token Counting, Models, Files, Skills, and beta APIs for Agents, Sessions, and Environments. A Claude Console account plus an API key or workload identity federation rule is required. The documentation also publishes request-size limits and explains that rate limits and spend limits are organized into usage tiers. Those operational limits matter for a production integration, particularly when an agent's workflow uses files or managed sessions.

## Pricing is usage based and varies by model and token direction

Anthropic's [pricing documentation](https://platform.claude.com/docs/en/about-claude/pricing) lists model input, cache-write, cache-read, and output rates in US dollars per million tokens. The published entry rate shown in the table is $0.10 per million input tokens for Claude Haiku 5.5 prompts up to 100,000 tokens, while output tokens and longer prompts have separate rates. It is therefore not a flat subscription price for API use. The same page says batch processing receives a 50 percent discount on input and output tokens. Prompt caching has distinct write and read multipliers, and some server-side tools can add usage charges beyond normal input and output token billing. Claude also has a [standing Free plan](https://claude.com/pricing) for its chat product, separate from the API's metered model pricing.

## MCP support centers on connectors and developer-created servers

On the Claude product side, a [connector](https://claude.com/docs/connectors/getting-started) links Claude to an outside app or service so it can find information and take actions. Claude users can add a connector from Anthropic's directory, use it within a conversation, manage its permissions, or disconnect it. The connector documentation says web Claude supports remote connectors, including MCP Apps, while the desktop app can also use local desktop extensions. For developers, Anthropic's [Python Claude Agent SDK](https://github.com/anthropics/claude-agent-sdk-python) documents custom tools implemented as in-process MCP servers inside a Python application. The documentation describes this as a way to expose application-defined tools to Claude, not as a single vendor-operated business-system connector.

## Published source is available for selected developer components

Anthropic maintains a verified [GitHub organization](https://github.com/anthropics) tied to anthropic.com. Its [Claude Agent SDK for Python repository](https://github.com/anthropics/claude-agent-sdk-python) is marked with the MIT License and documents installation as a Python package. The repository includes examples for tools, hooks, sessions, plugins, and an MCP calculator. This open-source availability applies to that developer component. The product documentation separately presents Claude itself as a hosted API and managed-agent service, so the SDK license should not be treated as a statement that every part of the hosted offering has the same license.

## The company frames its product work through a public-benefit mandate

Anthropic describes itself as a [public benefit corporation](https://www.anthropic.com/company) dedicated to securing AI's benefits and mitigating its risks. Its company page says the organization conducts research, develops safety techniques, and deploys resulting systems through partnerships and products. It names Claude as one of the practical tools intended for businesses, nonprofits, civil society groups, clients, and people around the world. The same page identifies a board that includes Dario Amodei and Daniela Amodei and describes the Long-Term Benefit Trust as part of its governance. These are company statements about mission and structure, not independent performance assessments of the models or service.

Our take

Check whether your deployment needs regional compliance support or a specific cloud path, since Anthropic routes through Claude.ai, Bedrock, Vertex AI, and Microsoft Foundry. Also verify that your workflow can use the 1M context window and browser/computer-use agents as intended.

## Anthropic price and terms

### Terms

- Free to start: Limited free use 50 free hours of usage daily per organization

### For AI agents

- MCP server: Yes [github.com, 29 October 2026](https://github.com/anthropics/claude-agent-sdk-python)

[![Anthropic](https://agentsindex.ai/img/1024/screenshots/anthropic/pricing.webp)](https://agentsindex.ai/img/1920/screenshots/anthropic/pricing.webp)

Pricing page, captured 19 October 2026

[anthropic.com/pricing](https://www.anthropic.com/pricing), checked 30 October 2026

## What we checked on Anthropic

Yes Partly No

How you build with it Yes

### MCP server

Anthropic's Python Agent SDK can package application-defined custom tools as in-process MCP servers.

> Custom tools are implemented in-process MCP servers that run directly within your Python application.

README, Custom Tools [github.com/anthropics/claude-agent-sdk-…](https://github.com/anthropics/claude-agent-sdk-python) Checked 29 October 2026

How you build with it Yes

### Open source

Anthropic publishes the Python Claude Agent SDK repository under the MIT License.

On the page: Repository header, license field

Repository header, license field [github.com/anthropics/claude-agent-sdk-…](https://github.com/anthropics/claude-agent-sdk-python) Checked 29 October 2026

## Anthropic alternatives

- [Fireworks AI](https://agentsindex.ai/fireworks-ai) · Inference, model routing and fine-tuning for open models on one OpenAI-compatible endpoint, serverless or in your cloud. · From $0.05 per 1m input tokens Checked 6 Nov · HIPAA SOC 2 · [Visit](https://agentsindex.ai/go/fireworks-ai)
  - MCP server: partly
  - API: yes
  - Open source: no
  - Self-host: partly
- Anthropic and Fireworks AI
  | Anthropic | From $0.10 per 1m input tokens | 4 Nov |
  | --- | --- | --- |
  | [Fireworks AI](https://agentsindex.ai/fireworks-ai) | From $0.05 per 1m input tokens | 6 Nov |

## Questions about Anthropic

- How much does Anthropic cost?: Anthropic costs $0.10 per 1m input tokens. From its own pages, checked 30 October 2026.
- Does Anthropic have an API or an MCP server?: Anthropic has an MCP server. From github.com/anthropics/claude-agent-sdk-python, checked 29 October 2026.
- Is Anthropic open source, and can you host it yourself?: Anthropic is open source. From github.com/anthropics/claude-agent-sdk-python, checked 29 October 2026.
- What have we checked on Anthropic?: 2 facts, under one heading: How you build with it.
- How do we check Anthropic?: We hold 2 facts on Anthropic, quoted from 2 pages. Facts last checked 29 October 2026. Price last checked 30 October 2026. Last change found 30 October 2026.

[How we check every listing](https://agentsindex.ai/methodology)

## Sources and changes

Last changed 30 October 2026

Sources 2

- Read 29 October 2026 · [github.com](https://github.com/anthropics/claude-agent-sdk-python) /anthropics/claude-agent-sdk-python
- Read 4 November 2026 · [anthropic.com](https://www.anthropic.com/legal) /legal

What changed 1

- 30 October 2026 Free tier

## Anthropic price history

We have read the price once so far, on 30 October 2026: from $0.10 per 1m input tokens.

| Price | Plan | First seen | Last seen | Readings | Source |
| --- | --- | --- | --- | --- | --- |
| $0.10 per 1m input tokens |   | 30 October 2026 | 30 October 2026 | 1 |   |

From our research on 30 October 2026. A date is when we read the page, not when the price changed.

## Similar to Anthropic

- [ERIUS PHONE](https://agentsindex.ai/eriusphone) Rented cloud Android phones, each in its own container, that keep an agent's apps and sign-ins between sessions. From about $44 a month
- [TinyFish](https://agentsindex.ai/tinyfish) Web search, page fetching, cloud browsers and goal-driven web automation for AI agents, with search and fetch free. Free plan, then $0.002 per minute
- [Smithery](https://agentsindex.ai/smithery) Public MCP server registry, now run by Arcade, listing direct vendor MCP URLs after its proxy shut down. Free plan
- [UI Discovery](https://agentsindex.ai/uidiscovery) A library of real websites, iOS apps and dashboards with each one's palette, fonts and original build prompt. From $20 a month

## From our research

[All research](https://agentsindex.ai/blog)

- [Comparison](https://agentsindex.ai/blog/daytona-vs-e2b-vs-modal) 16 Oct 2026 Daytona vs E2B vs Modal: which sandbox for agent code? Daytona, E2B and Modal compared on sandbox pricing, concurrency, GPU support and open-source status, plus why Daytona's flagship repo is now archived.
- [Buyer's guide](https://agentsindex.ai/blog/erius-phone-ai-agent-android-device) 9 Oct 2026 ERIUS PHONE review (2026): pricing, pros and cons ERIUS PHONE rents AI agents real cloud Android phones they drive by API, priced monthly per phone in euros, but it offers no webhooks.
- [Buyer's guide](https://agentsindex.ai/blog/hyper-review-autonomous-coding-agent) 13 Oct 2026 Hyper review (2026): an autonomous coding agent that ships verified pull requests Hyper plans, tests and delivers pull requests to your GitHub repository from an isolated cloud sandbox, with a verification check on each one and a hard spending cap on every run.

## Is Anthropic yours?

Free, and the page then carries the verified owner mark. You confirm it with an email at anthropic.com. [Claim this listing](https://agentsindex.ai/claim/anthropic?from=owner-line)

[Report an error](mailto:hello@agentsindex.ai?subject=Report%20an%20error%3A%20Anthropic%20(anthropic)&body=What%20is%20wrong%20on%20the%20page%2C%20and%20where%20on%20the%20product) if a fact is wrong.
