Anthropic has launched Claude Sonnet 5.5 as a faster and more efficient successor to Sonnet 5 without increasing its listed token prices. The model is aimed at everyday coding, bug fixes, document creation, and other well-scoped work rather than the most demanding tasks reserved for Opus.
In its Claude Sonnet 5.5 announcement, Anthropic says the model generates output more than 30 percent faster and can cost up to 30 percent less per completed task. Those performance and efficiency figures come from the company and should not be read as independent benchmarks for every workload.
Token prices stay at Sonnet 5 levels
Anthropic lists Sonnet 5.5 at $2 per million input tokens and $10 per million output tokens. Cache reads cost $0.20 per million tokens, while cache writes cost $2.50 per million. The claimed savings come from using fewer tokens to finish a task, not from a lower per-token rate.
Don’t miss the best of The Mac Observer
Set us as a preferred source and our Apple reporting ranks higher in your Google Search results and Discover feed — one tap, no account changes.
That distinction matters for developers estimating an application’s bill. A task can be cheaper if the model reaches the result in fewer steps, but actual spending will still depend on prompt size, output length, cache use, tool calls, and the effort setting. Anthropic sets Medium effort as the default in Claude apps and Claude Code, while its developer platform defaults to High.
The company positions Sonnet 5.5 below Claude Opus 5.5, which costs $4 per million input tokens and $20 per million output tokens. Anthropic says Sonnet is strongest for routine, clearly bounded work, while Opus remains intended for problems that require more careful judgment.
Coding and safety claims need real-world checks
Anthropic reports a 70.6 percent result on Terminal-Bench 4.0 and says the model improves long-horizon work and image understanding. It is also the first Sonnet release to launch with the company’s higher-level cyber safeguards because Anthropic considers its cybersecurity capabilities comparable to those of Opus 5.
Mac users can already work with Claude through several development workflows. MacObserver has explained Claude Code’s iOS Simulator support and previously covered Claude models on iPhone and iPad. Teams considering Sonnet 5.5 should compare it with their current model using representative tasks, since a vendor benchmark cannot predict the speed, quality, or cost of a specific codebase.
Discussion