Anthropic unveils Claude Sonnet 4.6 with 1M-token context and stronger coding

Anthropic releases Claude Sonnet 4.6, its most capable Sonnet model yet, featuring a 1 million-token context window, improved safety with fewer hallucinations, and enhanced coding abilities. It’s accessible via claude.ai, Claude Cowork, and API across major cloud platforms, with free usage limits and a Pro plan at $20/month (or $17/month if billed annually). API pricing starts at $3 per million input tokens and $15 per million output tokens. In benchmarks, Sonnet 4.6 outperforms Gemini 3 Pro and GPT 5.2 on agentic financial analysis and office tasks and generally beats Opus 4.6 on many tasks, though Opus 4.6 scores higher on Humanity’s Last Exam; it’s also advertised as cheaper than Opus 4.6.
- Anthropic releases Claude Sonnet 4.6: Benchmark performance, how to try it Mashable
- Introducing Claude Sonnet 4.6 Anthropic
- Anthropic releases Claude Sonnet 4.6, continuing breakneck pace of AI model releases CNBC
- Anthropic Says New AI Model Is Better at Using Computers Bloomberg.com
- Anthropic releases Sonnet 4.6 TechCrunch
Reading Insights
0
4
2 min
vs 3 min read
81%
573 → 111 words
Want the full story? Read the original article
Read on Mashable