Introducing Claude Sonnet 4.5
What happened
Anthropic announced Claude Sonnet 4.5, a new AI model marketed as the best coding model and strongest for building complex agents and using computers. The model shows gains in reasoning, math, and software coding, topping benchmarks like SWE‑bench Verified and OSWorld with a 61.4% score, up from 42.2% for the prior version. It can maintain focus for over 30 hours on multi‑step tasks and includes new features such as checkpoints in Claude Code, a refreshed terminal, a VS Code extension, context editing, and a memory tool for longer agent runs. The Claude Agent SDK and Chrome extension let developers and users run code, create files, and interact with browsers directly. Pricing stays at $3/$15 per million tokens, and the model is available via the Claude API today.
Why it matters
For everyday readers, the improvements mean AI assistants will be better at handling complex, multi‑step digital work like spreadsheet automation or browsing tasks, potentially saving time and reducing errors. Developers and businesses could see faster software development cycles and more reliable AI‑driven tools.
Reality check
An automated, sceptical read of the sourceNotable. The release adds substantial performance jumps and new developer tools, but builds on an existing model line rather than introducing a completely new paradigm.
Is this a big deal?
Get stories like this every morning
The day's AI news in plain English, in your inbox at 8:30 IST. Free.
More from Anthropic
Anthropic Upgrades Claude Code Projects to Run Parallel AI Workers Hands-FreeAnthropic · 14 Nov 2026Claude designed FPGA scaler board for 6th-gen consolesAnthropic · 13 Nov 2026I asked Claude Fable 5.2 to turn my site into a 90s Silicon Graphics workstationAnthropic · 13 Nov 2026Anthropic is endorsing SB 53Anthropic · 13 Nov 2026Understanding and addressing AI harmsAnthropic · 13 Nov 2026This page was collected from anthropic.com and summarised automatically by AIPulsee's pipeline. See how it runs.