AI Product Updates: New Features and Releases

Anthropic said it is increasing five-hour rate limits on Pro, Max and Team plans and giving subscribers a rate limit reset they can save and use whenever they choose.

Anthropic launched Opus 5.5, which it calls a noticeably better collaborator, with a link to more details.

The new model puts key information up front and follows the writing rules users give it, making long sessions easier to follow and addressing common feedback on Opus 5.

At its default effort setting, Opus 5.5 reaches frontier-level results at lower cost per task and generates output more than 30% faster than Opus 5.

Because Opus 5.5 needs less compute to serve, tests show it costs about 40% less than Opus 5 on typical workloads at default settings.

The release is described as a major step up from Opus 5, leading on agentic coding, computer use, and knowledge work.

Before release, the model was tested by external evaluators including METR and Frontier Design, and it achieves the strongest score to date on the most comprehensive alignment test.

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

An unsourced post claims Opus 5.5 leads across agentic coding, reasoning, computer use and visual recognition, at half the price of Fable 5.1.

Recruiting AI company Metaview integrated typesafe's jev into every agent, cutting candidate searches from minutes to seconds with the same accuracy and lower cost per search.

Verona's latest mainnet upgrade moves the x/zk Barretenberg verifier to v5.2.0, letting Noir and bb proofs verify at the protocol level without a self-maintained verifier contract.

Linear has launched MCP support for its Inbox, letting agents read and manage your Linear notifications.