Skip to content
AnthropicAnthropic

Claude Coded: Sonnet 4.5, Claude Code 2.0, and more.

We're covering everything new with Claude for developers, including the launch of Claude Sonnet 4.5, major updates to Claude Code, powerful new API capabilities, and exciting features in the Claude app. Join the Claude Developer Discord - https://anthropic.com/discord Check out the Claude Code Docs - https://docs.claude.com/en/docs/claude-code/overview Check out the Memory Cookbook - https://github.com/anthropics/claude-cookbooks/blob/main/tool_use/memory_cookbook.ipynb 00:06 Claude Sonnet 4.5 - Leading on SWE-bench and can focus for 30+ hours on complex tasks 01:03 Claude Code VS Code Extension - Now available in beta on the VS Code Marketplace 01:30 Claude Code 2.0 - Refreshed terminal UI 01:45 Claude Code Checkpoints - Roll back to previous states with /rewind 02:23 Usage Tracking - Monitor your usage in real-time with /usage command 02:35 Claude API - Context editing and memory tools for building sophisticated agents 03:37 Claude Agent SDK - Build your own agents with the tools that power Claude Code 03:59 Claude File Creation - Generate Excel, PowerPoint, Word, and PDF files in the Claude app

Oct 5, 20254mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Anthropic updates: Sonnet 4.5, Claude Code 2.0, agent tools

  1. Claude Sonnet 4.5 is positioned as Anthropic’s top coding model, leading SWE-bench Verified at 77.2% and showing long-horizon task focus beyond 30 hours.
  2. Claude Code now has a native VS Code extension (beta) with a sidebar and inline diffs, bringing Claude-driven edits directly into the IDE workflow.
  3. Claude Code 2.0 refreshes the terminal experience with better UI/status visibility, searchable prompt history, and a new checkpoint/rollback system via /rewind or double-Escape.
  4. The Claude API introduces context editing to automatically prune stale tool calls near token limits and a file-based memory tool to persist information client-side across conversations.
  5. Anthropic renamed and expanded the Claude Agent SDK (formerly Claude Code SDK) and added Claude app file generation for common formats like Excel, PowerPoint, Word, and PDFs for paid plans (preview).

IDEAS WORTH REMEMBERING

5 ideas

Sonnet 4.5 targets elite coding and long-horizon execution.

Anthropic highlights a 77.2% SWE-bench Verified score and claims the model can stay focused on complex tasks for 30+ hours, signaling improved endurance for multi-step engineering work.

Computer-use capability is becoming a first-class benchmark for Claude.

The transcript cites OSWorld improvement from 42% to 61% in four months and points users to the Claude for Chrome extension as a practical way to experience these gains.

Claude Code is being integrated into mainstream IDE workflows.

The new VS Code extension (beta) brings Claude Code into the IDE with real-time visibility into edits via a dedicated sidebar and inline diffs, reducing reliance on terminal-only operation.

Terminal users get meaningful usability upgrades in Claude Code 2.0.

A refreshed UI, improved status visibility, and searchable prompt history aim to make iterative prompting and long sessions easier to manage from the command line.

Checkpoints reduce the risk of letting an agent run large changes.

Using /rewind or double-Escape, developers can restore code, conversation, or both to a prior checkpoint, but the rollback applies only to Claude-made edits (not user edits or bash commands), so version control is still recommended.

WORDS WORTH SAVING

5 quotes

Claude Sonnet 4.5 is now available wherever you get your Claude, and it is the best coding model in the world.

Unknown

It is leading on SWE-bench Verified with a score of 77.2%, and we have actually seen it stay focused on complex tasks for well over 30 hours straight.

Unknown

But the feature that I am most excited about is a new checkpoints feature that lets you confidently run large tasks and roll back instantly to a previous state if needed.

Unknown

Context editing automatically clears stale tool calls and results from within the context window when approaching token limits.

Unknown

The memory tool, on the other hand, enables Claude to store and consult information outside of the context window through a file-based system.

Unknown

SWE-bench Verified performance and long-running focusReasoning/math/computer-use gains (OSWorld)Claude for Chrome extension expansionVS Code extension for Claude Code (beta)Claude Code 2.0 terminal UI and prompt historyCheckpoints and rollback with /rewindAPI context editing and file-based memory toolClaude Agent SDK for custom agentsClaude app: data analysis and document/file generation

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome