OpenAI and Anthropic New Model Releases (They Are At War)

OpenAI and Anthropic are engaged in an intense battle, simultaneously releasing new state-of-the-art coding AI models and launching competitive, and at times misleading, advertising campaigns.

• OpenAI and Anthropic released their new advanced coding AI models, GPT 5.3 Codeex and Claude Opus 4.6, within an hour of each other.

• Anthropic ran Super Bowl ads that controversially misrepresented how ads would appear in AI models, drawing a public response from OpenAI's CEO, Sam Altman.

• Both new models are focused on enhancing coding capabilities, with Claude Opus 4.6 offering a massive context window and GPT 5.3 Codeex demonstrating self-improvement capabilities.

A fierce competition has erupted between AI giants OpenAI and Anthropic, marked by the near-simultaneous release of their advanced coding models, GPT 5.3 Codeex and Claude Opus 4.6, on the same day. This rivalry extends to public advertising, with Anthropic running Super Bowl ads that implicitly criticize OpenAI's upcoming ad integration in ChatGPT. The ongoing competition is ultimately seen as beneficial for consumers, driving continuous innovation and ensuring accountability among AI providers.

AI Market Battle & User Base

• 00:00:00 An intense competition has emerged between AI companies Anthropic (makers of Claude) and OpenAI (makers of ChatGPT), likened to a 'David vs. Goliath' scenario due to OpenAI's significantly larger user base of 415 million monthly visitors compared to Claude's 15.5 million. Despite Claude's smaller market share, it is highly regarded within the AI community, particularly for coding, and the rivalry includes both new model releases and advertising campaigns.

Simultaneous Model Releases

• 00:01:13 OpenAI and Anthropic launched their most advanced AI models, GPT 5.3 Codeex and Claude Opus 4.6 respectively, less than an hour apart on February 5th, with Anthropic slightly pre-empting OpenAI. Both models are primarily designed for coders and aim to be state-of-the-art in agentic coding tasks. This close timing underscores the direct competition between the two companies for technological leadership.

Competitive Advertising & Controversy

• 00:02:11 The rivalry between OpenAI and Anthropic extended to advertising, notably during the Super Bowl, where Anthropic launched ads that misleadingly depicted AI responses interspersed with advertisements, appearing to critique OpenAI's announced ad integration in ChatGPT. Sam Altman, OpenAI's CEO, responded by calling Anthropic's ads 'dishonest,' clarifying that OpenAI's ads would be clearly labeled and external to chat responses. Altman also leveraged OpenAI's larger user base to justify their need for monetization through ads to maintain accessibility, while the attention generated by his rebuttal significantly outpaced the ad's original views.

Model Capabilities & Performance

• 00:08:14 Claude Opus 4.6 features a 1 million token context window, enabling it to process roughly 750,000 words for advanced coding, financial analysis, and research, with enhanced multitasking and adaptive thinking. GPT 5.3 Codeex is touted as the most capable agentic coding model to date, uniquely using early versions for self-debugging and development, indicating a significant acceleration in AI self-improvement. While benchmarks vary and are hard to compare directly, GPT 5.3 Codeex performed better on the Terminal Bench 2.0 for coding, while Claude Opus 4.6 showed superiority in agentic computer use, with a side-by-side test building a landing page yielding clean results from both, but with a slight edge to OpenAI's output quality.