Product · mentioned 309 times
Large language model family from Anthropic in its Claude line, positioned as the most capable top tier and used for advanced reasoning and coding.
Tap a name to go there · tap a line to see why they are linked
Startup Insider· 8 Nov
Philipp Müller says only OpenAI's Astra and Anthropic's Claude Faible and Claude Opus 5.5 can do this, and that the MIT authors named it weight-space programming.Latent Space: The AI Engineer Podcast· 7 Nov
Usage of Anthropic models other than Opus 5.5 fell by half or more within two to three months.MTS· 6 Nov
Bernardo Aceituno says hundreds of open-weight models perform better than Claude Opus 4.6 and can run on a local computer.The Information's TITV· 6 Nov
Dan Shipper says Claude is much better at capturing a particular person's style or voice, while Gemini is best at structuring and laying out a series of thoughts.etn.· 6 Nov
Anish says Anthropic distils Mythos into smaller models like Opus and Fable, which it releases, so it does not have to redo its pre-training runs.