Reviews
What the released models are really like, from tests and people who use them.
Model reviews
Near-Astra performance at Sonnet 5.5’s price.
Read →GoogleGemini 4 ArgonLevel with GPT-6 Astra in the first independent tests, behind Opus 5.5.
Read →AnthropicFable 5.1What’s new, what it costs and where you can use it.
Read →All modelsNine models, side by sidePick a route for the job, not a winner for everyone.
Compare →Hands-on
What people found after using them.
The excitement is about personality and taste, not just getting more done.
@mckaywrigley ↗On Opus 5.5A C-to-Rust portHAProxy moved with Opus 5.5 and Fable 5.1. Nearly all tests passed.
@bcherny, Anthropic ↗Opus 5.5 vs AstraBetter overall, slower on some jobsPrefers Opus 5.5, but finished some tasks faster with Astra and Sol.
@bridgemindai ↗Opus 5.5 vs AstraA review that got challengedAstra xhigh disputed parts of an Opus 5.5 high review on two projects.
@leo114119 ↗How long Opus 5.5 on high lasts
From our own paid accounts. A bigger plan buys more room, not a better rate.
Compare the plans →We don’t sell Claude plans. Anthropic does.
Put it to work
Agent Skills your model can run today.
