Anthropic Releases Sonnet 5
Claude Sonnet 5 launches with lower API pricing, agentic capabilities, and scientific tools. Anthropic addresses account bans and high costs head on.
“AI Disruption” Publication 10100 Subscriptions 20% Discount Offer Link.
Claude has two long-standing issues: one is its tendency to ban accounts, and the other is its high cost.
The first issue has become very familiar to many Claude users in the past week. Discussions about Anthropic’s risk control measures on social platforms have grown increasingly bizarre — from IP bans and location-based login reviews to warnings like “don’t casually open emails from Anthropic,” because opening an email might reveal your location and be used to determine your account’s region.
While some rumored details still need verification, the frequent account bans are an indisputable fact. Yesterday, when I woke up, I discovered that my fifth Claude account had also been suspended.
As for the second issue — the high price — Anthropic has just given its latest answer.
Claude Sonnet 5 has officially launched. According to the official announcement, Sonnet 5 is currently the most agent-capable Sonnet model. It can create plans, call tools such as browsers and terminals, and autonomously execute complex tasks.
From a product positioning perspective, Anthropic is moving more capabilities that previously belonged to the high-end Opus models down to the more commonly used and more affordable Sonnet line. This upgrade isn’t just about making the model stronger — it’s also about rebalancing the relationship between “usability” and “affordability.”
