Claude Opus 5.5 cuts Opus prices and raises usage limits
On 22 September, Anthropic released Claude Opus 5.5, the first model in its Claude 5.5 family. Anthropic says it performs at the level of Claude Fable 5.1 on most work and costs 40% less to run than Opus 5 on typical workloads.
For people who pay by the token, the price list is the news. Opus 5.5 costs $4 per million input tokens and $20 per million output tokens, 20% less than Opus 5. Cache reads, the cheaper rate charged when the model rereads context it has already processed, drop 60% to $0.20 per million tokens. Anthropic says cache reads make up most of the cost of agent and coding work, which is why the overall saving is bigger than the headline price cut.
What subscribers get
Anthropic says it raised the five-hour usage limits on the Pro, Max, Team and seat-based Enterprise plans. Subscribers also get a rate limit reset that they can save and use whenever they choose. Anthropic says Opus 5.5 also generates output more than 30% faster than Opus 5. A fast mode in Claude Code and the Claude Platform runs up to 2.5 times faster at $8 per million input tokens and $40 per million output tokens.
How Anthropic compares it
In Anthropic's table, Opus 5.5 scores 66.4% on Terminal-Bench 4.0, a test of multi-step work in a command line, against 55.8% for Fable 5.1 and 57.9% for OpenAI's GPT-6 Astra, the last figure as reported by OpenAI. Anthropic adds its own caveat that at this level "benchmark margins have become a less reliable guide to real-world differences," and says that in its own use the gap with Fable 5.1 is narrower than the scores suggest.
On safety, Anthropic says the model was tested before release by outside evaluators including METR, and scored best of any model so far on its automated behavioral audit, a suite of thousands of simulated scenarios. Because it is close to Claude Mythos 5.1 in biology and cybersecurity, it ships with safeguards similar to Fable 5.1's.
What to do this week
If you run Opus 5 through the API, look at last month's bill and find the share that comes from cache reads. That share is where most of the saving shows up. Switch a test copy of one workload to claude-opus-5-5compare cost per task for a week, and check that the answers still pass whatever review you already do. If you use a Claude subscription, check whether your plan is one of those with raised limits.
What we think
A price cut together with a better model is the release most teams can act on right away. Anthropic is also candid that benchmark gaps at this level say less than they used to. We would treat the 40% figure as Anthropic's estimate on its own workloads and measure it on ours before changing any budget.
Sources