Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

We've acquired omniroute.online ! Try the #1 open source router and stop paying routing fees!

We’re ending our partnership with Cursor following its acquisition by SpaceX. Under our proposal, Cursor’s direct access to our models would end on November 12. We know that the people most affected by this decision are the developers who rely on OpenAI models in Cursor. We care about their experience in this transition and we’re ready to go above and beyond to support them. openai.com/index/our-decisio…

Prediction: Deepseek v4 flash will take over Claude as the no.1 winner in market share. It's the biggest story of 2026! Three reasons: 1. It's so cheap. 100x cheaper in unit token pricing. 2. It's so fast. Much much faster than Claude. My vibe check is around 2-3x faster. 3. It's so cache-efficient: most of my tokens spent are in cache read, and cache input pricing are $0.5/Mt Opus 5 v.s. $0.0028/Mt Deepseek. Deepseek has a much higher cache hit rate, which means the overall effective cost per task could be upwards of 500x cheaper. For something that's 500x cheaper AND 2-3x faster per task, why wouldn't Deepseek V4 Flash win.