Getting started (API keys, console)
The Anthropic Developer Console (console.anthropic.com) is the web-based administrative platform that gives developers everything they need to start using the Claude API. From the Console, you can generate and manage API keys, add team members, configure billing and prepaid credits, view usage analytics, and experiment with models in the built-in Workbench before writing any code.
API keys are the authentication tokens that identify your organization's requests to the Claude API. Every API call must include a valid key in the request header. Keys follow the format 'sk-ant-...' and are shown only once at creation time, so they must be saved immediately in a secure location.
Importantly, Claude API access is completely separate from claude.ai subscriptions (Free, Pro, Max, Team). A Pro or Max subscription gives you enhanced chat features on claude.ai but does not include API access. API usage is billed separately on a pay-as-you-go basis tied to token consumption, and you must add prepaid credits to your Console account before making API calls.
When you’d use it
- ◆Rapid Prototyping in the Workbench — A developer wants to test different system prompts and model behaviors before writing any application code. They use the Console's built-in Workbench to interactively send messages, adjust parameters, and evaluate responses, then copy the working configuration into their codebase.
- ◆Environment-Segregated Key Management — An engineering team runs separate development, staging, and production environments. They create a distinct API key for each environment, named clearly (e.g., 'backend-prod', 'backend-staging'). If a key is compromised, only that environment is affected and the key can be revoked without disrupting others.
- ◆Customer-Facing Chatbot Integration — A SaaS company embeds Claude into their support portal. They generate a dedicated API key for the support service, set it as a server-side environment variable, and call the Messages API to respond to user questions — keeping the key out of client-side code entirely.
- ◆Cost-Optimized Batch Document Processing — A data team needs to classify thousands of support tickets. Instead of making individual real-time API calls, they use the Batch API to submit all requests at once, achieving a 50% cost reduction on both input and output tokens compared to standard synchronous calls.
- ◆Usage Monitoring and Budget Control — A startup wants to avoid surprise overage bills. They use the Console's usage dashboard to monitor daily token consumption per key, set spend alerts, and understand which services are driving costs before scaling up.
What changed recently
- ◆2025-11-24 — Claude Opus 4.5 released at $5 per million input tokens and $25 per million output tokens, making Opus-level capabilities more broadly accessible.
- ◆2026-02-05 — Claude Opus 4.8 released as the most capable Anthropic model to date, priced at $5 per million input tokens and $25 per million output tokens, with up to 90% savings via prompt caching and 50% savings via batch processing.
- ◆2026-02-17 — Claude Sonnet 4.6 released at $3 per million input tokens and $15 per million output tokens, with up to 90% cost savings with prompt caching and 50% cost savings with batch processing.
- ◆2026-03-16 — Anthropic updated official support documentation to explicitly clarify that paid claude.ai subscriptions do not include Claude API or Console access — these are separate products with separate billing.
This is the short version
The full chapter has three worked examples, the common pitfalls, and the workflow that makes it pay — plus the other 84 features, kept current.
Get Claude Master — $97 →