Claude New Features Getting Started Guide: Extended Output, Dashboard, and Release Notes

Claude has been updating at a rapid pace lately, and the most practical changes are concentrated in model capabilities, output length, and visualization/management on the developer side. Below, in the order of “things you can use right away,” I’ll break down Claude’s new features clearly and provide ready-to-follow usage tips.

Claude Sonnet 3.5: How to use a faster workhorse model

In Anthropic’s developer newsletter, Claude Sonnet 3.5 is positioned as a new version that is “more speed- and cost-friendly, while also performing strongly in evaluations.” In practice, it’s better suited for everyday tasks like high-frequency conversations, summarization and rewriting, and code explanation—and you can try it directly using the Claude Console workbench.

If you often ran into “it stops halfway because it isn’t long enough” in Claude before, prioritize the extended output feature below—it’s an immediate, noticeable upgrade.

Extended Output: Claude can write more in a single go (most crucial on the API side)

The Claude API’s maximum output limit for Sonnet 3.5 has been increased from 4096 to 8192, but it must be explicitly enabled. To do this, add the request header: "anthropic-beta": "max-tokens-3-5-sonnet-2024-07-15", and set max_tokens to the range you need.

This is especially helpful for long-form outlines, batch generation of manuals, and outputting multiple blocks of code at once; questions that previously had to be split across multiple rounds can now be written more smoothly in one shot in Claude.

Usage and Cost Dashboard: Claude billing is finally easy to check

The developer console now includes Usage and Cost tabs, letting you track consumption by dollar amount, token count, and API key. For scenarios where a team shares a key or a single project runs multiple scripts, this dashboard helps quickly pinpoint “who exactly blew up the Claude usage.”

It’s recommended to manage your commonly used keys with distinct names, first review the weekly trend in the dashboard, and then decide whether to adopt more aggressive prompt compression or caching strategies.

Release Notes and Documentation Upgrades: Fewer pitfalls matter more than more features

Claude’s documentation now includes more complete release notes covering updates to the API, the Claude Console, and the Claude app. For developers, this is far more reliable than “digging through community posts everywhere,” especially when investigating why the output style of the same piece of code suddenly changed.

In addition, the official Claude Cookbook has been expanded to fill in key guides on citations, RAG (retrieval-augmented generation), and classification. If you’re building knowledge-base Q&A or doing content moderation, using these patterns can help you get Claude tuned into a stable, reusable state more quickly.