Anthropic has launched Claude Sonnet 4.5, aiming to outdo other artificial intelligence (AI)-powered models in coding, help in complex agents, and manage extended computer tasks.
The platform stated that Sonnet 4.5 primarily focused on multi-step projects for over 30 hours and ranked the highest in the benchmarks, such as SWE-bench Verified and OSWorld at 61.4%.
Alongside the launch, Anthropic released product upgrades, including checkpoints for Claude Code, a new context editing feature, a memory feature in the Claude API, and the ability to implement code and generate files directly within Claude apps.
Additionally, the firm has released the Claude Agent SDK, allowing developers to build AI agents using its infrastructure and introduce a Chrome extension for Max users.
Anthropic strongly advised Sonnet 4.5 as a sharp drop-in replacement across apps, APIs, and Claude Code.
Meanwhile, OpenAI reported $4.3 billion in revenue for the first half of 2025, marking a significant 16% year-over-year increase, as per The Information.
The platform’s cash burn skyrocketed to $2.5 billion, primarily from research and development, with $6.7 billion spent in that area.
Notably, Anthropic’s rival, OpenAI, held $17.5 billion in cash and security midyear and is aiming for $13 billion in full-year revenue with an $8.5 billion burn rate.
To boost monetisation, OpenAI recently released “Instant Checkout” on ChatGPT, allowing US users to buy items from Etsy and Shopify sellers directly within chats.
