2 min read
Anthropic’s new Claude Sonnet 4.6 promises Opus-level coding at Sonnet pricing
Anthropic launched Claude Sonnet 4.6, a new AI model offering near-flagship Opus 4.6 performance at a lower cost, excelling in coding and office tasks.
3 articles
Anthropic launched Claude Sonnet 4.6, a new AI model offering near-flagship Opus 4.6 performance at a lower cost, excelling in coding and office tasks.
A developer built Context Lens to trace LLM context usage. Testing four AI coding tools on the same bug fix revealed vastly different token consumption and strategies: Claude Opus was surgical (23K tokens), Codex efficient (29-47K), Claude Sonnet balanced (42-44K), and Gemini profligate (up to 350K tokens). The tools show no deliberate context management; efficiency stems from investigation strategy, not optimization.
Paper on using LLMs for theorem proving to verify the seL4 microkernel, aiming for industrial-scale real-world application.