Anthropic releases Claude Sonnet 5.5 at 30% faster, 30% cheaper
On September 28, 2026, Anthropic released Claude Sonnet 5.5, designed for everyday business tasks like bug fixing, document generation, and spreadsheet work. The model keeps Sonnet 5's pricing at $2 per million input tokens and $10 per million output tokens, but runs more than 30% faster and costs up to 30% less per task through more efficient token usage. Sonnet 5.5 complements Claude Opus 5.5, which handles complex work requiring careful judgment.
In short
- Claude Sonnet 5.5 runs 30% faster and costs 30% less per task while keeping the same $2/$10 per-million-token pricing as Sonnet 5
- Cost savings come from fewer tokens and tool calls per task, not from API rate cuts, making the upgrade valuable for high-volume automation
- Designed for everyday business tasks: document generation, bug fixing, spreadsheets, and data processing
- Available on AWS, Google Cloud, and Microsoft Azure with zero data retention
- Claude Haiku 5.5 is coming soon for even higher-volume, cost-sensitive workloads
What happened
On September 28, 2026, Anthropic announced Claude Sonnet 5.5, the second model in the Claude 5.5 family. According to Anthropic, the model runs more than 30% faster than Claude Sonnet 5 and costs up to 30% less per task, despite keeping the same API pricing of $2 per million input tokens and $10 per million output tokens.
The cost reduction does not come from lower API rates. Instead, Anthropic found in testing that Sonnet 5.5 uses fewer tokens and fewer tool calls to complete the same work. This efficiency gain means the total cost per task drops even though the per-token rate stays the same.
Why it matters for businesses that automate work
Automation platforms that rely on LLM APIs see their costs tied to token consumption. A 30% reduction in tokens per task directly cuts the cost of running the same workflows. This matters most for businesses that run high-volume, repeating automation tasks: data entry, follow-ups, reconciliations, report generation, document handling, and billing work.
Speed also affects automation value. When a model responds 30% faster, multi-step workflows finish sooner. For tasks that involve user feedback loops, faster iterations mean quicker turnaround and less idle time waiting for API responses.
Anthropic positioned Sonnet 5.5 for "well-scoped everyday tasks, fixing bugs, and creating polished documents, slides, and spreadsheets." These are exactly the kinds of routine work that automation targets. The model is available on all major cloud platforms—Amazon Web Services, Google Cloud, and Microsoft Azure—under the model ID claude-sonnet-5-5.
What changes in practice
For a business that has built automation workflows using Claude Sonnet 5, swapping the model ID to Sonnet 5.5 is straightforward. The API interface is the same. What changes behind the scenes is efficiency: the same prompt produces output faster and consumes fewer tokens.
In real workflows, this means lower monthly API bills without renegotiating terms. A business running 10,000 document-generation tasks per month on Sonnet 5 could reduce token spend by roughly 30% by switching to Sonnet 5.5, assuming the same prompt and task scope.
One practical note: Anthropic says users who run Sonnet 5 with thinking disabled must switch to the new "between_tools" setting before moving to Sonnet 5.5. This is a breaking change for existing deployments, but the setting is documented in Anthropic's migration guides.
Security and safeguards
Sonnet 5.5 is the first Sonnet model to launch with cybersecurity safeguards comparable to Opus 5. According to Anthropic, it has stronger protections against high-risk requests than Sonnet 5. Biology safeguards remain unchanged. The model also includes invisible text watermarking to help detect AI-generated content, added to comply with regulations such as the European Union's AI Act.
What to watch next
Anthropic said Claude Haiku 5.5, a smaller model designed for high-volume and cost-sensitive applications, will join the Claude 5.5 family in the coming weeks. Haiku 5.5 is expected to offer even lower costs per task for workloads that do not require Sonnet's reasoning depth.
Businesses building or expanding automation should track Haiku 5.5's release and pricing. For purely routine tasks like data extraction from templates or simple form filling, Haiku may offer better cost-per-token economics than Sonnet.
Anthropic also released Opus 5.5 on September 22, 2026, one week before Sonnet. Opus remains the flagship model for complex reasoning and judgment-heavy work. The staggered releases of Opus and Sonnet within the same family suggest Anthropic is competing on speed and cost efficiency across price tiers, not just raw capability.
Sources
How AiStaffo would automate this
AiStaffo designs automation that handles routine back-office work: data entry, follow-ups, reconciliations, reports, billing, and document handling. Claude Sonnet 5.5's 30% speed and token-efficiency gains directly reduce the API costs of running these workflows at scale. When a business automates 5,000 document generations or data-entry tasks per month, a 30% reduction in token consumption per task cuts costs without changing the work the automation does. AiStaffo would integrate Sonnet 5.5 into existing automation pipelines, update the model ID in deployment, and pass the efficiency savings to clients. Book a free automation audit to see which routine tasks in your business could run faster and cheaper on this model.
Questions people ask
Is Claude Sonnet 5.5 cheaper to use than Sonnet 5?
What kinds of tasks is Sonnet 5.5 best for?
Can I swap from Sonnet 5 to Sonnet 5.5 in existing code?
Where can I access Claude Sonnet 5.5?
Is Claude Haiku 5.5 available yet?
Book a free automation audit
Thirty minutes. We look at one process you run every week and tell you exactly what an AI worker would take off your desk, and what it would not.
