Claude Sonnet 5.5 is Faster, Smarter and Up to 30% Cheaper

Anthropic has introduced Claude Sonnet 5.5, its latest mid-tier AI model, with major gains in coding, knowledge work, computer apply, and visual understanding.

Key points

• Claude Opus 5.5 Beats Fable 5.1 at Much Lower Cost and Faster Speed.
• It additionally reaches 46.2% on FrontierCode 1.1 at maximum effort and 55.5% on CursorBench 4.0.
• Claude Sonnet 5.5 is now available throughout Anthropic's platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure.
• Anthropic notes Sonnet 5.5 is intended to complement Claude Opus 5.5, with Sonnet targeting faster, well-scoped work while Opus remains focused on more complex tasks.
• The model additionally improves computer-use performance and visual chart recognition, with scores of 80.1% and 61.6%, respectively, on Anthropic's noted evaluations.

For context, the new model generates responses more than 30% faster than Claude Sonnet 5 and can cost up to 30% less per task since it uses fewer tokens to complete the same work.

Pricing remains at $2 per million input tokens and $10 per million output tokens, with cache reads at $0.20 per million tokens.

Claude Sonnet 5.5 brings a substantial improvement in agentic coding. It scores 70.6% on Terminal-Bench 4.0, compared with 10.3% for Sonnet 5.

It additionally reaches 46.2% on FrontierCode 1.1 at maximum effort and 55.5% on CursorBench 4.0. Anthropic notes the model can understand substantial codebases rapidly and complete coding tasks in fewer steps by batching tool calls more efficiently.

For context, the firm states Sonnet 5.5 can handle tens of thousands of lines of code and sustain multi-hour coding tasks while keeping responses fast.

Although Anthropic notes Opus 5.5 remains stronger for complex, open-ended work requiring sustained judgment, on some evaluations, its performance at higher effort levels approaches Claude Opus 5.5.

Anthropic positions Sonnet 5.5 as a model for well-defined everyday tasks that require a balance of speed, capability, and cost.

It is designed for tasks such as fixing bugs, creating documents, building presentations, and working with spreadsheets. The model additionally improves at after visual and design requirements, including creating polished user interfaces and presentations that need less manual editing.

On the GDPval-AA benchmark for real-world professional tasks throughout 44 occupations and nine industries, Sonnet 5.5 scores 1, 844, almost matching Opus 5.5 at 1, 846 and well above Sonnet 5's 1, 449.

For context, the model additionally improves computer-use performance and visual chart recognition, with scores of 80.1% and 61.6%, respectively, on Anthropic's noted evaluations. OpenAI Pauses Model Training After Agent Bypassed Restrictions While Training. It Can Beat Pokémon Red. Sonnet 5.5 additionally demonstrates stronger long-horizon reasoning and image understanding.

Anthropic notes it is the first Sonnet model capable of beating Pokémon Red while working only from screenshots. The result highlights its ability to interpret visual information and carry out tasks over extended sequences rather than simply responding to individual prompts. Lower Cost Per Task.

Although Sonnet 5.5 has the same listed token rates as Sonnet 5, Anthropic states the new model generally requires fewer tokens to complete tasks.

Notably, the firm reports that this makes Sonnet 5.5 up to 30% cheaper per task than its predecessor. At lower effort settings, Anthropic notes the model can additionally outperform Sonnet 5's previous best results on some benchmarks at a fraction of the cost. Its faster generation further reduces the time required for routine tasks. Sonnet 5.5 additionally introduces stronger safeguards alongside its improved capabilities.

Anthropic notes the model's cybersecurity abilities have improved significantly compared with Sonnet 5. Consequently, Sonnet 5.5 is the first Sonnet model to launch with cybersecurity safeguards and fallbacks similar to those applied for Anthropic's more capable models.

While higher-risk cybersecurity requests can fall back to Sonnet 5, routine software development and bug fixing remain backed.

For context, the model retains the same biology safeguards as Sonnet 5. Anthropic additionally states Sonnet 5.5 introduces extra protections against attempts to extract its reasoning through large-scale distillation attacks.

Claude Sonnet 5.5 is now available throughout Anthropic's platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure. Pricing Claude Sonnet 5.5. Input tokens $2 per million. Output tokens $10 per million. Cache reads $0.20 per million. Cache writes $2.50 per million.

Anthropic additionally stated Claude Haiku 5.5, designed for high-volume and cost-sensitive applications, will join the Claude 5.5 family in the coming weeks. Stay Connected with ProPakistani.

Obtain the latest tech news, telecom insights, and product launches wherever you prefer. Follow on Google Discover.

Add as a preferredSource on Google Follow on Google News Join WhatsApp.

Add ProPakistani to Preferred Sources and see more of our stories in Google Search and Top Stories.

Taken together, the developments around claude Sonnet 5.5 is Faster, Smarter and Up to 30% Cheaper point to a situation that is still moving, and the coming days should bring more clarity.