OpenAI and Anthropic go to war: Claude Opus 4.6 vs GPT 5.3 Codex
OpenAI launches GPT-5.3-Codex with token efficiency and GB200-NVL72 co-design as Anthropic releases Claude Opus 4.6 with agentic coding capabilities, with benchmarks citing 2.93× faster inference.
OpenAI launched GPT-5.3-Codex, emphasizing token efficiency, inference speed, and hardware/software co-design with GB200-NVL72 and NVIDIA collaboration. The new Frontier agent platform supports business-context agents with execution environments and learning capabilities. Anthropic showcased Opus 4.6 agent teams autonomously building a clean-room C compiler booting Linux, highlighting advances in agentic coding and long-context capabilities.
Community benchmarks report 2.93× faster inference and significant efficiency gains, signaling a shift away from infinite compute budgets in 2026.