US EditionPublished • loading... • Updated
Anthropic Says Its Newest AI Model Is Getting Pretty Good at Using a Computer
Claude Sonnet 4.6 matches Opus-level performance on key tasks while maintaining Sonnet pricing, enabling enterprises to run powerful AI agents more affordably, Anthropic said.
- On Tuesday, Anthropic rolled out Claude Sonnet 4.6 with a beta 1M token context window, making it the default for Free and Pro users across Claude interfaces and major platforms.
- Facing intense competition from OpenAI and Google, Anthropic positions Sonnet 4.6 as near-Opus intelligence at mid-tier cost and follows Opus 4.6 launched about twelve days earlier.
- Benchmark testing shows Sonnet 4.6 scored 79.6% on SWE-bench Verified and hit 63.3% on agentic financial analysis, outperforming Opus 4.6's 60.1%.
- The release immediately reshapes enterprise economics by making continuous agent deployments affordable, as Sonnet 4.6 largely eliminates the cost-quality trade-off for enterprises processing 10 million tokens per day.
- Anthropic is positioning Sonnet 4.6 for regulated industries as Infosys, Indian IT giant, partners to integrate Claude into its Topaz AI platform, while the company opened its first India office in Bengaluru and India accounts for about 6% of Claude usage.
Insights by Ground AI
Podcasts & Opinions
23 Articles
23 Articles
Coverage Details
Total News Sources23
Leaning Left1Leaning Right0Center9Last UpdatedBias Distribution90% Center
Bias Distribution
- 90% of the sources are Center
90% Center
C 90%
Factuality
To view factuality data please Upgrade to Premium
