Anthropic又“踢馆”！Sonnet 4.6操作电脑接近人类，性能堪比旗舰模型、定价仅1/5

Core Insights - Anthropic has launched Claude Sonnet 4.6, a new AI model that offers flagship-level performance at a mid-range price, significantly altering the pricing landscape in the AI industry [1][10] - The model's pricing remains the same as its predecessor Sonnet 4.5, at $3 per million tokens for input and $15 for output, while the flagship Opus model is priced five times higher [10][11] - The release comes amid Anthropic's aggressive push into the enterprise market, highlighted by a recent $30 billion funding round that doubled its valuation to $380 billion [2] Performance Enhancements - Claude Sonnet 4.6 has shown a fivefold improvement in computer operation capabilities over 16 months, achieving a score of 72.5% on the OSWorld benchmark, nearing human-level performance [3][5] - In programming tasks, developers preferred Sonnet 4.6 over Sonnet 4.5 in approximately 70% of cases, and it outperformed the flagship Opus 4.5 in 59% of scenarios [7][8] - The model's performance in various benchmarks is competitive with Opus 4.6, scoring 79.6% in SWE-bench Verified and 72.5% in OSWorld-Verified tests [8][9] Cost-Effectiveness - The cost-performance ratio of Sonnet 4.6 is transformative for enterprises making millions of API calls daily, eliminating the need to choose between lower-quality results and high-cost options [10][11] - Early testers reported that Sonnet 4.6's performance matched or exceeded that of the more expensive Opus models, making it a clear choice for many organizations [12][11] Strategic Capabilities - Sonnet 4.6 features a 1 million token context window, allowing it to handle extensive documents and perform long-term strategic planning effectively [12][13] - The model demonstrated a unique ability to develop novel strategies in a simulated business environment, significantly outperforming its predecessor in profitability [13][14] Competitive Landscape - The rapid release of Sonnet 4.6 reflects the intense competition in the AI industry, with Anthropic launching significant updates within a short timeframe [16] - Concerns have arisen among investors regarding the potential disruption of traditional software companies by AI advancements, as evidenced by recent stock market reactions [17][16] - Sonnet 4.6 has outperformed competitors like Google’s Gemini 3 Pro and OpenAI’s GPT-5.2 in several benchmarks, indicating its strong position in the market [19][20]