GPT-5 Codex vs Claude Code: free attack on Anthropic
OpenAI introduced GPT-5 Codex. A specialized version of their flagship model, completely reimagined for programming!
The new model replaces the outdated o3 in the Codex web client and is already available in local tools Codex CLI and plugin for popular IDEs. The access policy is interesting. The service is free for all ChatGPT subscribers regardless of plan. Essentially, GPT-5 Codex is a direct attack on competing solution Claude Code from Anthropic.
The model underwent specialized fine-tuning on a large dataset of real tasks. From creating projects from scratch to large-scale refactoring and code reviews. Although improvement on standard SWE-bench Verified benchmark seems insignificant. 74.5% versus previous 72.8%. But OpenAI says their internal tests show impressive jump in refactoring tasks. From 33.9% to 51.3%. We have to take their word for it.
The most interesting achievement is adaptive distribution of computational resources. The model generates shorter responses for simple tasks and more detailed ones for complex ones. According to OpenAI, GPT-5 Codex also demonstrates unprecedented autonomy, working up to 7 hours straight on complex tasks. Iteratively improving solutions and fixing errors in tests.
AIvengo >
Reviews >
GPT-5 Codex vs Claude Code: free attack on Anthropic
Почитать из последнего
UBTech will send Walker S2 robots to serve on China's border for $37 million
Chinese company UBTech won a contract for $37 million. And will send humanoid robots Walker S2 to serve on China's border with Vietnam. South China Morning Post reports that the robots will interact with tourists and staff, perform logistics operations, inspect cargo and patrol the area. And characteristically — they can independently change their battery.
AI chatbots generate content that exacerbates eating disorders
A joint study by Stanford University and the Center for Democracy and Technology showed a disturbing picture. Chatbots with artificial intelligence pose a serious risk to people with eating disorders. Scientists warn that neural networks hand out harmful advice about diets. They suggest ways to hide the disorder and generate "inspiring weight loss content" that worsens the problem.
OpenAGI released the Lux model that overtakes Google and OpenAI
Startup OpenAGI released the Lux model for computer control and claims this is a breakthrough. According to benchmarks, the model overtakes analogues from Google, OpenAI and Anthropic by a whole generation. Moreover, it works faster. About 1 second per step instead of 3 seconds for competitors. And 10 times cheaper in cost per processing 1 token.
