AI models / The AI briefing

Anthropic introduces Claude Opus 5.5 with lower API pricing

Anthropic’s model update also changes pricing and usage. Comparing the whole workload matters more than a headline benchmark.

Anthropic introduced Claude Opus 5.5, announcing lower input, output and cache-read token prices than Opus 5. The company reports improvements in coding, knowledge work and communication, alongside stronger results in its safety evaluations. It also announced higher five-hour usage limits for Pro, Max, Team and seat-based Enterprise subscriptions.

What Anthropic is changing

Anthropic introduces Claude Opus 5.5 with API pricing changes covering input, output and cached tokens compared with Opus 5. The announcement also describes higher usage allowances for eligible subscription and enterprise plans.

The company reports benchmark results and discusses safety evaluations. These are claims in the developer’s announcement; they should be distinguished from independent testing of a particular business application.

What a useful comparison includes

A team considering a model change can compare response quality, latency and total cost on its own representative tasks. Token prices are only one part of that calculation: retries, output length, cache use and the time spent reviewing results also influence the outcome.

Original source

This report summarises the source below. Analysis is labelled separately; product and research claims remain attributed to their source.

Read the original at Anthropic

← Back to all AI news