OpenAI Unveils GPT-6.2 Astra and Sol Models Ahead of Anthropic IPO
OpenAI unexpectedly released the GPT-6.2 Astra and Sol models just before Anthropic's IPO, delivering superior performance in professional workflows, scientific research, and coding while significantly cutting costs and token usage. This strategic launch right before a major competitor's IPO highlights the intense competition in the AI sector, demonstrating that frontier models can achieve higher performance benchmarks at a fraction of their predecessors' operating costs. GPT-6.2 Astra surpasses Claude Fable 5.2 on professional benchmarks like Agents' Last Exam while costing only 5% as much, and its input/output API prices are 20% lower than the previous generation. Meanwhile, GPT-6.2 Sol targets everyday efficient tasks with pricing identical to GPT-6.1 Sol.
## BACKGROUND
Anthropic is a major artificial intelligence safety and research company known for developing the Claude series of AI assistants, preparing for a public stock offering. Benchmarks like Agents' Last Exam and Terminal-Bench-Science evaluate AI agents on complex, multi-step real-world professional and scientific tasks.