Anthropic's Claude AI Takes Benchmark Lead Amid Rising Competition

Anthropic’s Claude model recently topped a major AI benchmark, overtaking OpenAI’s GPT‑5.5 in categories such as advanced reasoning, multi‑step problem solving, language understanding and coding challenges. The ranking shift highlights intensified rivalry in the generative‑AI market, where major players including Anthropic, OpenAI, Google DeepMind and Chinese firm Z. AI are rapidly advancing capabilities.

At the same time, developers can now access Claude‑compatible APIs at significantly reduced rates, with pricing as low as $0.20 per million input tokens for Claude 3 Haiku and $1.00 per million for Claude 3 Opus. The lower costs are intended to improve profitability for AI start‑ups, SaaS founders and freelancers building chatbots, coding assistants and automation tools.

Parallel to Claude’s rise, Chinese lab Z. AI released GLM‑5.2, an open‑source model that rivals Claude Opus 4.8 and GPT‑5.5 on benchmarks such as FrontierSWE and SWE‑bench Pro. GLM‑5.2 runs on Huawei Ascend chips without Nvidia hardware and is offered under an MIT license, with token pricing notably cheaper than Claude’s.

Educational offerings have also emerged, featuring e‑degree courses and professional training on using Claude for marketing, prompt engineering and workflow automation, priced as low as $19.99.