The Claude-lash is here: Opus 4.7 is burning through tokens — and some people's patience

Anthropic released a new model, Opus 4.7. Some users on X and Reddit aren't happy, calling it "dumb" and more costly on compute than previous models.
Hosted on MSN

Claude Code creator shares Opus 4.7 setup tips

Boris Cherny, creator of Claude Code, has detailed how to optimize the new Opus 4.7 model for faster, safer, and more efficient workflows. His guidance covers using Auto mode to reduce permission prompts, leveraging recap and focus modes for cleaner ...

Anthropic’s Claude Is Pumping Out Vulnerable Code, Cyber Experts Warn

Anthropic’s latest Claude models are introducing serious security issues into code, cyber experts say. The company is yet to officially explain why.
The Droid Guy

Did Anthropic Secretly Nerf Claude? The Evidence, the Bugs, and What Actually Happened

Over the past month, thousands of developers noticed something wrong with Claude. Responses felt dumber. Token limits ran out in minutes instead of
Infosecurity Magazine

Anthropic Rolls Out Claude Security for AI Vulnerability Scanning

Claude Security enters public beta, giving enterprises AI driven code scanning with no API integration or custom agents required
Decrypt

DeepClaude Lets You Run Claude Code With DeepSeek's Brain for 17x Cheaper

DeepClaude swaps Claude Code's expensive Anthropic backend for DeepSeek V4 Pro, keeping the agent loop, slashing the bill.