Cohere open-sources a coding agent that runs on a single H100

Cohere's North Mini Code ranks 8th of 127 open-weight models on output speed — but generates 3x the output tokens of ...
i-SCOOP

Kimi K2.7 Code, the open-weight coding model that thinks 30% less

A deep dive into Kimi K2.7 Code from Moonshot AI: architecture, benchmarks, pricing, and how to put its 256K context and ...

AI Coding Agents Write 180% More Code But Ship Only 30% More Software

AI coding agents boost code output by 180% but shipping rises only 30%, MIT finds. Why private data access beats benchmark ...
腾讯网

MiniMax低调发布M3和MiniMax Code，我们测完后有几点感受

前沿的 Coding 能力、1M 的上下文窗口，还有原生的多模态 作者: GaKi 编辑: Koji‍ 排版: NCon就在今天，MiniMax 发布了新一代模型 M3。这次最值得看的地方，是 M3 终于在同一个模型里凑齐了三种能力：前沿的 Coding 能力、1M 的上下文窗口，还有原生的多模态。同一时间，MiniMax 还更新了配套的 Agent 产品 ...

"Claude Code is writing all the code": Anthropic's output is up 8x, creator says

I haven't written a line of code by hand in 8 months.” At Fortune Brainstorm Tech, Anthropic Head of Claude Code Boris Cherny ...
腾讯网

Claude Code 史诗级更新！AI 编程进入“反向教学”模式，写到一半会 ...

先说 Claude Code：现在支持用 /output-style 来定制沟通风格，Output styles允许你把 Claude Code 调整为任意类型的Agent来配合你的工作流程，同时保留它的核心能力（运行本地脚本、读写文件、维护 TODO） 这次主打两种风格，专为学生和开发者提升技能而来： ...
CIO

AI coding tools are changing output faster than they are changing judgment

AI is like a super-fast junior dev: it’s great at drafting code quickly, but you still need a human brain to spot the risky logic and big-picture flaws. I say this not as a spectator to the AI tooling ...
InfoWorld

Output from vibe coding tools prone to critical security flaws, study finds

Popular vibe coding platforms consistently generate insecure code in response to common programming prompts, including creating vulnerabilities rated as ‘critical,’ new testing has found. Security ...