Anthropic is reportedly evaluating safety and considering the release of a new Claude model amid competitive pressure from OpenAI's GPT-6 Astra, which is…
1 editorial report · 2 verified social mentions. The most authoritative report leads while later evidence completes the story.
Last time I counted how many actions it takes to get to the first prompt with OpenSpec, Spec Kit, BMAD and Kiro. The usual reply was that I measured unboxing, not mowing the lawn. Fair. So this time I built a real project first: TypeScript, Express, SQLite with migrations, React on the front, one repo, real git history, built with plain Claude Code and no workflow at all. Then each tool got an identical copy and the same one sentence: let me drag a task from one day column to another. Before the runs I wrote down which files a correct implementation has to touch and which ones it must not. The trap was that the server already handled the feature, the PATCH route and the gateway both take day_index. Anything that planned a migration or a new endpoint would be inventing work that was already
Open mention🤖 Anthropic may already be preparing another Claude. Reuters says Anthropic is considering a new model as GPT-6 Astra starts putting pressure on Claude. Astra is already accounting for 13% of tracked enterprise AI spending, versus 8% for Claude Fable. Reuters also reports that Anthropic is evaluating the safety of its next model as it weighs a potential release. Just days ago, Dario Amodei argued that frontier AI development needs to slow down. Now Anthropic may have to accelerate again simply to keep pace with OpenAI. @aipost 🏴
Open mention