Anthropic shipped Claude 4.5 Opus this morning with a 1M-token context window, native computer-use APIs, and a third pricing tier aimed at enterprise customers running long-running agent workloads.
The big-context push lands the same week Gemini 2.5 Flash matched it at 2M and OpenAI signaled GPT-5 Turbo will follow. For builders, the practical upshot is that you can now feed entire codebases, multi-hour transcripts, or quarter-long policy archives without resorting to RAG plumbing — at least when latency and cost are not the binding constraints.
Pricing for the new Enterprise tier was not disclosed at announcement; partners we spoke with described it as 'meaningful' but not punitive, and Anthropic confirmed the standard Opus rates remain unchanged at $15/$75 per million tokens.
We will publish a head-to-head benchmark against GPT-5 and Gemini 2.5 Pro by end of week.