Anthropic is rolling out a new AI workhorse that promises to run powerful autonomous agents without the price shocks or cyber risks that have dogged top-tier models.

Anthropic sets the stage: agentic power at lower cost

On June 30, Anthropic formally introduced Claude Sonnet 5, describing it as “the most agentic Sonnet model yet,” able to make plans, use browsers and terminals, and run autonomously at levels that recently required larger, more expensive systems.¹ The company says Sonnet 5 narrows the gap with its flagship Opus 4.8 while remaining cheaper and showing “a much lower ability to perform cybersecurity tasks” than Opus in safety tests.¹

Anthropic positioned Sonnet 5 as a clear step up from its predecessor Sonnet 4.6 across reasoning, tool use, coding, and knowledge work, while reporting “an overall lower rate of undesirable behaviors” in agentic contexts.¹

Media frame the launch as a cost and safety play

Technology outlets quickly cast the release as part of a broader shift toward agentic AI becoming table stakes. TechCrunch noted that Sonnet 5 “promises performance close to that of Opus 4.8, but for much lower costs,” arguing that the real differentiator now is “how cheaply” and “how reliably without human oversight” models can run agents.²

The Next Web stressed pricing, reporting that Sonnet 5 “runs close to the flagship Opus 4.8 on many tasks, but costs less than half as much,” with introductory rates of $2 per million input tokens and $10 per million output tokens, rising to $3 and $15 after August 31, while Opus 4.8 is priced at $5 and $25.³ However, it highlighted a catch: a new tokenizer can increase token counts by up to 1.35x, muting the apparent savings.³

Axios underscored the regulatory and security backdrop, calling Sonnet 5 “a lower-priced model designed to bring more agentic AI capabilities to everyday users without the same cyber-risk profile as its most powerful systems,” and noting it “was not deliberately trained on cybersecurity tasks” and has a “much lower ability” to perform dangerous cyber activities than current Opus models.⁴

Market reaction: deployment as an orchestrator

On launch day, Perplexity AI’s CEO Aravind Srinivas signaled immediate adoption, announcing on X: “New orchestrator model for Computer users (Pro and Max): Sonnet 5.”Aravind Srinivas (AravSrinivas)⁵ In a repost of Perplexity’s own update, he amplified that “Claude Sonnet 5 is now available in Perplexity for Pro and Max subscribers” and can be selected “as an orchestrator model in Computer,” indicating early demand for cheaper, capable agent backends.Aravind Srinivas (AravSrinivas)⁶

Across Anthropic’s own messaging and outside coverage, the throughline is clear: Sonnet 5 is a bid to make always-on AI agents economically and politically sustainable—by trading a slice of raw capability for significant gains in affordability and cyber safety.

---

[1] "Introducing Claude Sonnet 5" – Anthropic describes Sonnet 5’s agentic abilities, near‑Opus performance, and reduced cybersecurity capabilities compared to Opus.

[2] "Anthropic launches Claude Sonnet 5 as a cheaper way to run agents" – TechCrunch frames Sonnet 5 as confirmation that agentic capability is now baseline, with cost and reliability becoming key differentiators.

[3] "Anthropic launches Claude Sonnet 5, a cheaper agent model" – The Next Web details Sonnet 5’s pricing versus Opus 4.8 and explains how a new tokenizer complicates headline cost savings.

[4] "Anthropic debuts Claude Sonnet 5 for everyday agent tasks with lower cyber risk" – Axios highlights Sonnet 5’s lower cyber-risk profile and notes it wasn’t deliberately trained on cybersecurity tasks.

[5] @AravSrinivas on X – "New orchestrator model for Computer users (Pro and Max): Sonnet 5."

[6] @AravSrinivas on X – "Claude Sonnet 5 is now available in Perplexity for Pro and Max subscribers. You can also select it as an orchestrator model in Computer. https://t.co/UktzCrUZU6"