A legal team reviewing a large contract set
- Upload long contracts into a single request using the 1M-token context
- Ask claude-opus-5-5 to list risky clauses
- Cache the shared instructions to reduce repeat cost
needs API key
Anthropic · API platform
Claude Developer Platform (Claude API)
Developer access to Anthropic's Claude models with usage-based pricing, batch discounts and prompt caching.
- Provider
- Anthropic
- Latest confirmed model
- Claude Fable 5.1 (claude-fable-5-1); Claude Opus 5.5 is the recommended default
- Pricing
- Paid only
- Free plan
- No
- API
- Yes · free quota
- Status
- Generally available
- Last verified
- Oct 19, 2026
What is Claude Developer Platform (Claude API)?
The Claude Developer Platform gives programmers API access to Anthropic's Claude models, plus a browser console for testing prompts. Current models are Claude Fable 5.1 for demanding reasoning, Claude Opus 5.5 (the suggested starting point), Claude Sonnet 5.5 and Claude Haiku 5.5 for fast, high-volume work. All current models accept text and images, support tool use and have a 1M-token context window.
Building assistants, agents and document workflows on a hosted model without managing GPUs or model training.
Why use Claude Developer Platform (Claude API)?
Practical benefits
- 1M-token context window on all current models at standard pricing
- 50% Batch API discount
- Deep prompt-caching discounts (cache reads 2.5-10% of input price)
- Available directly and through Amazon Bedrock, Google Cloud and Microsoft Foundry
Who it suits
- Developers building agents and coding tools
- Enterprises processing long documents
- Teams needing a choice of cloud platforms
When another tool may fit better
- Non-technical users who just want to chat (use the Claude app)
- Offline or self-hosted deployments
- Projects needing a permanent free API allowance
- Learning curve
- Some experience helps
- Skills needed
- Programming and API basics; prompt design.
Example workflows
- Prototype a prompt in the Claude Console, then call claude-opus-5-5 from your code
- Route simple classification to claude-haiku-5-5 and hard tasks to claude-fable-5-1
What can it do?
- Four-tier model lineupFable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5, plus legacy models still available.
- Adaptive thinking and effort controlModels decide how much to reason, steered by an effort parameter.
- Message Batches API50% off input and output tokens for asynchronous jobs; up to 300K output tokens with a beta header on supported models.
- Prompt caching5-minute and 1-hour cache writes with discounted cache reads.
- Server toolsWeb search, web fetch and code execution tools; code execution is free when used with recent web search/fetch tool versions.
- Claude Mythos 5.1 LimitedSame price as Fable 5.1 but offered with limited availability.
Good for
- Agentic coding tools
- Long-document analysis and Q&A
- Customer support assistants
- Data extraction and classification at scale
Where it runs
Pricing and plans
Paid only Usage-based per 1M tokens. Fable 5.1 is $10 input / $50 output, Opus 5.5 $4 / $20, Sonnet 5.5 $2 / $10 and Haiku 5.5 from $0.10 / $0.50 (higher above 100K-token prompts). Batch requests are 50% off and cache reads cost a small fraction of input price.
| Plan | Price | Billing | Free trial / tier | Usage limits | Main inclusions | Model access | API access | Restrictions |
|---|---|---|---|---|---|---|---|---|
| Pay as you go (Standard) | Billed per 1M input/output tokens by model, with separate cache-write and cache-hit rates. | usage-based | New users receive a small amount of free credits to test the API (amount not stated). | Rate limits by usage tier |
|
Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5 and legacy models | Yes | Mythos models have limited availability |
| Batch | 50% discount on input and output tokens. | usage-based | — | Asynchronous processing |
|
Current models | Yes | Fast mode not available with Batch |
| Enterprise | Contact provider | usage-based | Extended trials for enterprise evaluation via sales | — | — | — | Yes | — |
Worked example (usage-based API pricing): sending 1,000,000 input tokens and receiving 200,000 output tokens at the listed claude-fable-5-1 rates ($10 and $50 per 1M) costs about $20 before tax. Real bills depend on the model, caching, batch discounts and any tier you qualify for.
API and developer access
- Official API
- Yes
- Documentation
- API docs (opens in a new tab)
- Authentication
- API key (x-api-key header)
- Pricing model
- per 1M tokens
- Free API quota
- New users receive a small amount of free credits to test the API; amount not specified.
- Rate limits
- Tier-based
- Context / input limits
- 1M tokens on all current models; 128K max output (synchronous)
- SDKs
- Not documented
- Streaming
- Not documented
- Batch processing
- Yes
- Structured output
- Not documented
- Tool / function calling
- Yes
- Commercial use
- See the provider terms
- Data retention & privacy
- See the provider privacy terms
Published API rates
| Item | Price | Unit |
|---|---|---|
| claude-fable-5-1 input | $10 | per 1M input tokens |
| claude-fable-5-1 output | $50 | per 1M output tokens |
| claude-opus-5-5 input | $4 | per 1M input tokens |
| claude-opus-5-5 output | $20 | per 1M output tokens |
| claude-sonnet-5-5 input | $2 | per 1M input tokens |
| claude-sonnet-5-5 output | $10 | per 1M output tokens |
| claude-haiku-5-5 input (prompts up to 100K tokens) | $0.10 | per 1M input tokens |
| claude-haiku-5-5 output (prompts up to 100K tokens) | $0.50 | per 1M output tokens |
| claude-haiku-5-5 input (prompts over 100K tokens) | $0.50 | per 1M input tokens |
| claude-haiku-5-5 output (prompts over 100K tokens) | $2.50 | per 1M output tokens |
A free Claude Developer Platform (Claude API) account is not the same thing as free API access. The API allowance above is the provider's documented developer quota.
Advantages and disadvantages
Advantages
- 1M context on all current models without a long-context surcharge (except Haiku 5.5 pricing by prompt length)
- Strong caching and batch discounts
- Multi-cloud availability with matching model IDs
Disadvantages
- Top-tier Fable 5.1 costs $50 per 1M output tokens
- Haiku 5.5 price rises for prompts over 100,000 tokens
- Only a small, unspecified amount of free credit for new users
Privacy: Retention and data-use terms not verified in this research; see Anthropic's commercial terms.
Real-world examples
A marketplace classifying product listings
- Send listing text to claude-haiku-5-5 with category rules
- Use the Batch API for the back catalogue at 50% off
- Write categories back to the database
needs API key
A developer testing prompts before coding
- Sign in to the Claude Console
- Draft and compare prompts in the browser
- Copy the generated code snippet into the app
needs API key
Claude Developer Platform (Claude API) alternatives
OpenAI API
Pay-per-token developer access to OpenAI's GPT models for building apps, agents and automations.
Claude Developer Platform (Claude API) vs OpenAI APIGemini API and Google AI Studio
Google's developer API for Gemini models, with a free tier in Google AI Studio and paid tiers for production.
Claude Developer Platform (Claude API) vs Gemini API and Google AI StudioOpenRouter
One OpenAI-compatible API key for hundreds of models from many providers.
Claude Developer Platform (Claude API) vs OpenRouterTogether AI
Cloud platform for running, fine-tuning and training open models on serverless APIs or rented GPUs.
Claude Developer Platform (Claude API) vs Together AICompare Claude Developer Platform (Claude API) with these side by side
Frequently asked questions
Is the Claude API free?
Not on an ongoing basis. Anthropic says new users get a small amount of free credits to test the API; after that, usage is billed per token.
Which Claude model should I start with?
Anthropic's docs suggest Claude Opus 5.5 for most workloads, Claude Fable 5.1 for the most demanding reasoning, and Claude Haiku 5.5 for high-volume, latency-sensitive tasks.
How much does Claude Opus 5.5 cost?
$4 per 1M input tokens and $20 per 1M output tokens on the Claude API, with Batch requests at 50% off.
What is the context window?
All current models (Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5) have a 1M-token context window and up to 128K output tokens synchronously.
Is the API different from a Claude.ai subscription?
Yes. Claude.ai plans cover the consumer app; API usage is billed separately through the Claude Developer Platform.
Can I use Claude through my cloud provider?
Yes. The same models are offered on Amazon Bedrock, Google Cloud, Microsoft Foundry and Claude Platform on AWS, each with its own pricing and lifecycle details.