Claude Sonnet 4.6: Benchmark performance, how to try it

Anthropic has just released its latest Large Language Model (LLM), Claude Sonnett 4.6. The Tuesday release quickly follows the launch of Claude Opus 4.6, the company's premium AI model, on Feb. 5.

According to Anthropic, "Claude Sonnet 4.6 is our most capable Sonnet model yet." The company says Sonnet 4.6 has a 1 million token context window in beta. Crucially, Anthropic reports that Sonnet 4.6 performed well on internal safety tests, showing a low tendency to hallucinate and engage in sycophancy. 

"Sonnet 4.6 brings much-improved coding skills to more of our users," Anthropic said, referring to Claude's popularity among developers who use AI to code.

If you're looking to use Anthropic's latest AI model, the company has made it really easy. Here's how to access Clause Sonnet 4.6.

For both free and Pro users, Claude Sonnett 4.6 is available now as the default model on claude.ai and Claude Cowork. Anthropic has also rolled the model out through its API and all major cloud platforms.

Free users will have limited usage rates that depend on current demand. Limits reset every five hours. For those who need higher limits, Claude Sonnet 4.6 costs the same price rate as the previous model. The Claude Pro plan costs $20 per month or $17 per month if paid annual. If going through the API, Claude Sonnett 4.6 starts at $3 per million input tokens and $15 per million output tokens.

Claude Sonnet 4.6 benchmark performance

According to Anthropic's benchmark tests, Claude Sonnet 4.6 is the company's most powerful model for agentic financial analysis and office tasks, beating out competitors like Google's Gemini 3 Pro and OpenAI's GPT 5.2. 

On those tasks, Claude Sonnet 4.6 also beats out Anthropic's own Opus 4.6, Anthropic's most powerful AI model. 

In its release announcement, Anthropic said that many developers with early access to Claude Sonnet 4.6 preferred the model — not just to its predecessor, Claude Sonnet 4.5, but also Claude Opus 4.5. According to the Sonnet 4.6 system card, the new model improves on key benchmarks like Humanity's Last Exam, though Claude Opus 4.6 scored higher.

Benchmark performance

• GPQA Diamond: 89.9 percent

• ARC-AGI-2: 58.3 percent

• MMMLU: 89.3 percent

• SWE-bench Verified: 79.6 percent

• HLE (Humanity's Last Exam): With tools 49.0 percent, without tools 33.2 percent

AI-powered insurance company Pace told VentureBeat that Sonnet 4.6 scored the best out of any Claude model on its complex insurance computer use benchmark.

These results are notable as Claude Opus models are generally the more intelligent and preferable for complex reasoning.

Claude Sonnet 4.6 is not only more powerful than some Opus models, but more affordable too. As previously mentioned, Claude Sonnet 4.6 is priced at $3/$15, whereas Opus 4.6's rates are $5/$25.