#

Key takeaways

• Anthropic announced Claude Sonnet 5.5 on September 28, 2026.

• It is the second model in the Claude 5.5 family and complements Claude Opus 5.5.

• Anthropic says it is built for well-scoped everyday work, bug fixing, and documents.

• The company reports faster output, lower per-task cost, and new safeguards.

• The announcement gives no Morocco-specific availability information.

Claude Sonnet 5.5 arrives as a new mid-tier model

Anthropic announced Claude Sonnet 5.5 on September 28, 2026. The company calls it the second model in the Claude 5.5 family. It also presents the model as a complement to Claude Opus 5.5.

Anthropic positions Sonnet 5.5 for well-scoped everyday work. It also names bug fixing, polished documents, slides, and spreadsheets as target tasks. In the company's framing, Opus is better for complex work that needs sustained judgment.

Reported benchmark results

Anthropic shared several benchmark results for Sonnet 5.5. It reports a score of 70.6% on Terminal-Bench 4.0. The company compares that with 10.3% for Sonnet 5.

Anthropic also reports 46.2% at Max effort on FrontierCode 1.1 Main. It says Sonnet 5.5 is two points below Opus 5.5 on GDPval-AA. These are company-reported benchmark results, and Anthropic cautions that benchmark scores show only one part of a model's capabilities.

Pricing, speed, and efficiency claims

Anthropic lists Sonnet 5.5 at $2 per million input tokens and $10 per million output tokens. Those are the same list prices as Sonnet 5. The company says the new model usually uses fewer tokens, which can make it up to 30% less expensive per task in its tests.

Anthropic also says Sonnet 5.5 generates output more than 30% faster than Sonnet 5. These are Anthropic's reported claims, not independent findings. The announcement does not provide outside verification for the pricing or performance comparisons.

Safety and governance notes

Anthropic says Sonnet 5.5 includes alignment, cybersecurity, and biology safeguards. It also describes visible fallback to Sonnet 5 for certain higher-risk cybersecurity tasks. That suggests the company is using model routing as part of its safety approach.

The source does not give detailed technical thresholds for those safeguards. It also does not explain how the fallback works in every case. Readers should treat the safety description as part of Anthropic's announcement, not as a full technical audit.

Availability across platforms

Anthropic says Sonnet 5.5 is available on all platforms. The announcement names Anthropic's own service, Amazon Web Services, Google Cloud, and Microsoft Azure. The source does not add any Morocco-specific availability information.

That means the announcement supports a general global reading only. If a reader is evaluating deployment, the practical lesson is to confirm access and pricing directly on the relevant platform.

Morocco relevance

The source reports no Morocco-specific availability information. The only safe takeaway is conditional and global: readers should verify platform access, pricing, and task fit before planning adoption.

What this announcement means

The main message is straightforward. Anthropic is presenting Sonnet 5.5 as a faster and more efficient model for everyday work. It also places the model below Opus 5.5 on complexity, while still highlighting strong benchmark results.

Because the claims come from Anthropic, they should be read as vendor-reported performance. The announcement is useful for understanding product positioning, pricing, and safety framing. It is less useful as a standalone proof of real-world performance across every workload.