By Ken Muthomi - Reporter
USA , July 1 , 2026 - Anthropic has launched Claude Sonnet 5, a new AI model engineered to perform more complex tasks autonomously while offering developers improved performance at a lower cost than its flagship Opus models.
The company, which described the model as the most agentic Sonnet model to date, has substantial improvements in key aspects of agentic performance like reasoning, tool use, coding, and knowledge work, compared to its predecessor, Claude Sonnet 4.6.
Anthropic has said that the model will have the capability of planning tasks, using tools such as web browsers and computer terminals, and working independently on complex workflows that previously required larger and more expensive AI models.
The model will be available across all plans, including as the default model for Free and Pro plans, and is available to Max, Team, and Enterprise users.
It is also available in Claude Code and on the Claude Platform, where it launches with introductory pricing of $2 per million input tokens and $10 per million output tokens through August 31, 2026.
Beginning in September, it will be priced at $3 per million input tokens and $15 per million output tokens.
Anthropic has said that feedback from early access customers has established that the new model is capable of completing multi-step tasks that previous versions often left unfinished. The testers also indicated that the model could verify its own work without being explicitly instructed to do so.
“For many developers, the agentic AI era began with Sonnet-class models: Claude Sonnet 3.5, 3.6, and 3.7 were the first models that showed impressive skills in coding and tool use. More recently, though, the clearest gains in agentic capabilities have been in our Opus-class models,” the company said.
More from Kenya
“Sonnet 5 narrows the gap: its performance is close to that of Opus 4.8, but at lower prices. It’s a substantial improvement over its predecessor, Sonnet 4.6, on important aspects of agentic performance like reasoning, tool use, coding, and knowledge work,” it added.
Compared to the previous models, Sonnet 5 has also been equipped with better safety characteristics, which include refusing to reject malicious requests, resisting prompt injection attacks, and reducing hallucinations and sycophantic responses compared to Claude Sonnet 4.6.
Even though the model can perform routine cybersecurity-related tasks, it was not specifically trained for offensive cybersecurity work and performs substantially worse than its Opus models when tested on developing software exploits.
Anthropic has, however, enabled cybersecurity safeguards by default, allowing the system to detect and block potentially dangerous cyber-related requests in real time, since the model is notably more capable than its predecessors.
“Sonnet 5 was never able to develop a full working exploit, but it does show a slightly higher rate of partial success than Sonnet 4.6. This latter change is likely due to improvements in general intelligence rather than specific training,” the company stated.
Discover more Kenya updates