Najnowsze wiadomości i analizy Claude Opus | Yellow

Independent testers scored Anthropic's new Claude Opus 5 at 159 on one capability index, two points behind Claude Fable 5, while dividing sharply over its code review record. Key Points: Epoch AI rated Claude Opus 5 at 159 on its capability index, against 161 for Claude Fable 5, with the two tied on

Anthropic released Claude Opus 5 on Friday at $5 per million input tokens, saying it nears the performance of the pricier Claude Fable 5 at half the cost. Key Points: Opus 5 holds the same token prices as Opus 4.8, at $5 and $25 per million input and output tokens. The model scored 1,861 on the GDPv

Elon Musk said SpaceX will open its Grok 4.5 model to the public Thursday, calling the 1.5 trillion-parameter system an Opus-class rival that is faster and cheaper. Key Points: Musk plans to release Grok 4.5 to the public Thursday, faster and cheaper than Claude Opus. The model runs on a 1.5 trillio

Anthropic's Claude Fable 5 loses its subscription-included access on Jul. 7, after which the model costs double Claude Opus 4.8, pushing many users back toward the cheaper option. Key Points: Fable 5 stays included on paid Anthropic plans only through Jul. 7, then shifts to metered usage credits. Cr

Anthropic released Claude Sonnet 5, a new AI model it says brings Sonnet-level systems closer to Claude Opus 4.8 in agent work. Key Points: Anthropic says Claude Sonnet 5 improves reasoning, tool use, programming and knowledge work over Sonnet 4.6. The model has lower listed rates than Opus 4.8, but