Anthropic has released Claude Opus 4.7, featuring a significant improvement in agentic coding tasks with a SWE-bench Pro score increase to 64.3% and enhanced high-resolution vision capabilities. OpenAI followed shortly after with GPT-5.5, codenamed "Spud," which boasts a leading 82.7% score on Terminal-Bench 2.0 and a dramatic improvement in long-context reasoning, doubling its MRCR v2 score to 74.0% on 1M tokens. These advancements signal a shift towards more autonomous AI agent workflows in coding assistance, moving beyond simple chat-style interactions. AI
IMPACT These releases push AI coding tools towards autonomous agent workflows, signaling a rapid evolution beyond basic chat assistance.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on dev.to — Anthropic tag →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
