Anthropic Launches Claude Sonnet 5.5 with 30% Speed Gain and New Safety Features

A glowing geometric neural network node on a light gray background, symbolizing speed and intelligence in AI models.

Anthropic has launched Claude Sonnet 5.5, its latest mid-tier AI model, which the company says runs more than 30% faster than the previous generation while costing less to operate. The model targets everyday tasks such as document generation, summarization, and bug fixing, and is positioned as a practical upgrade for enterprise workflows.

Performance and Pricing

According to Anthropic, Sonnet 5.5 delivers significantly faster output speeds than its predecessor, making it the fastest Sonnet model to date. The model is priced identically to Sonnet 5 at $2 per million input tokens, $10 per million output tokens, and 20 cents per million cache reads. However, Anthropic claims the new model typically requires fewer tokens to complete the same work, resulting in an effective cost reduction of over 30% compared to the previous generation.

Capabilities and Testing

The model has been evaluated on several benchmarks. On Terminal-Bench 4.0, an agentic coding test, Sonnet 5.5 scored 70.6%, a sharp increase from the previous model’s 10.3%. It also scored within two points of the flagship Opus 5.5 on GDPval-AA, a test of real-world occupational tasks. Additionally, Sonnet 5.5 is the first in its class to beat Pokémon Red using only screenshots, a benchmark for long-horizon reasoning and visual understanding.

Safety and Compliance

Sonnet 5.5 includes invisible text watermarking designed to help detect AI-generated content. Anthropic states this feature does not affect text quality or readability and was added to comply with global regulations, including the European Union’s AI Act. The model also introduces cybersecurity safeguards inherited from Opus 5.5, triggering responses when prompts touch on sensitive topics like security or biology, though most software development work remains unaffected.

Enterprise Feedback and Future Releases

Early testers, including Zendesk, reported that Sonnet 5.5 resolved support tickets 20% faster and made fewer errors than previous Claude models in production. Anthropic also announced that Haiku 5.5, a smaller and more efficient model for high-volume applications, will follow in the coming weeks.

AI-generated editorial illustration; not a documentary photograph.

Source: Anthropic debuts Claude Sonnet 5.5 running 30% faster than the previous-generation AI model

Share this article
FacebookThreadsLinkedInX

Leave a Reply

Your email address will not be published. Required fields are marked *