
Anthropic launches Claude Sonnet 5 as a cheaper way to run agents
Anthropic has launched Claude Sonnet 5, positioning it as a powerful and affordable agentic model capable of autonomous planning and tool use.
The release underscores that agentic capability is becoming the baseline expectation across all price tiers, mirroring moves by OpenAI and Google.
Sonnet 5 is priced at $2 per million input tokens, making it cheaper than competitors like GPT-5.5 and Gemini 3.1 Pro.
Performance benchmarks show Sonnet 5 closely matching Opus 4.8 on reasoning and tool use, while slightly outperforming it on knowledge work tasks.
The model is available across all subscription plans, with free and Pro users getting it as the default starting Tuesday.
This shift signals that the competitive differentiator is now cost efficiency and reliability rather than just capability.
Early testers report that Sonnet 5 excels at completing complex, multi-step tasks without requiring explicit human oversight.


