Skip to main content
  1. Home
  2. Computing
  3. News

Claude’s Sonnet 5 is built to do more on its own and cost you less

Better than its predecessor, nearly as good as the flagship, and meaningfully cheaper than both.

Add as a preferred source on Google
Art, Floral Design, Graphics
Anthropic

Every major AI lab is racing to prove its models can work autonomously with minimal hand-holding; we’re now seeing pricing emerge as the next battleground. 

Anthropic just fired its latest shot, Claude Sonnet 5, a model the company says performs nearly as well as its flagship Opus 4.8 at a fraction of the cost.

Introducing Claude Sonnet 5, our most agentic Sonnet yet.

It makes plans, uses tools like browsers and terminals, and runs autonomously at a level that just a few months ago required larger and more expensive models. pic.twitter.com/UKK8G7ww5h

— Claude (@claudeai) June 30, 2026

So what’s actually new here?

Sonnet 5 is Anthropic’s most agentic Sonnet model yet. It can plan multi-step tasks, use tools like browsers and terminals, and complete work autonomously. Previously, doing that required a larger, more expensive model.

Recommended Videos

On one agentic coding benchmark, Sonnet 5 scores 63.2%, a meaningful jump over Sonnet 4.6’s 58.1%. However, it still trails Opus 4.8’s 69.2%. 

On knowledge-work tasks, though, Sonnet 5 slightly edges out Opus 4.8, but it is sort of given, especially since Opus is built for harder judgment calls. 

What does it cost, and is it actually safer?

Sonnet 5 launches today at $2 per million input tokens and $10 per million output tokens through August 31, 2026. After that, the prices increase to $3 and $15, respectively. 

That undercuts Opus 4.8, OpenAI’s GPT-5.5, and Google’s Gemini 3.1 Pro, though Gemini 3.5 Flash still remains cheaper. 

On the safety front, Anthropic reports Sonnet 5 hallucinates and shows sycophantic behavior less often than its predecessor. Furthermore, the AI model is notably weaker at dangerous cybersecurity tasks than Opus-class models, a deliberate tradeoff rather than an accident (via Anthropic). 

Anthropic’s Sonnet 5 is now available as the default model on the Free and Pro plans. It is accessible across Max, Team, Enterprise, Claude Code, and the API.

Sonnet 5’s launch follows a pattern set by rivals: OpenAI’s GPT-5.6 Sol entered preview just last week with subagent task-splitting, and Google’s Gemini 3.5 Flash, launched in May, is pitched explicitly as agentic rather than conversational.

Sonnet 5 also uses an updated tokenizer that can map the same input to up to 1.35x as many tokens as Sonnet 4.6, though introductory pricing is designed to offset that change.

Shikhar Mehrotra
For more than five years, Shikhar has consistently simplified developments in the field of consumer tech and presented them…
Google’s new Magic Pointer Play Store listing reveals a Gemini shortcut built for Googlebooks
The unannounced app turns the cursor into a contextual AI tool for search, image creation, and shopping
Plant, Text, Business Card

Google has quietly published a new Play Store listing for Magic Pointer, an unannounced app built for Googlebooks. Updated on July 10, the app turns the cursor into a Gemini shortcut that can act on whatever a user selects on screen.

Magic Pointer can send an image to Lens, generate a related image, or surface a shopping action without forcing users to open a separate chatbot. Regular Android devices currently show as incompatible, so the listing offers an early preview rather than a broad release.

Read more
You can stop using AI, but this new report says you probably can’t escape it
A UK survey found that most people feel AI exposure is unavoidable, raising harder questions about consent, privacy, and whether opting out is still realistic
AI Chatbots

More people are trying to use less AI, but avoiding it altogether may already be impossible.

A survey of 2,055 UK adults found that 42% deliberately limit how much AI they use. Another 70% said avoiding AI exposure would be difficult or impossible, even when they actively wanted less of it.

Read more
The face on an AI interviewer may matter as much as the decision it makes
Researchers found that race and gender matching changed how fairly rejected applicants viewed an automated interview, even though everyone received the same outcome
File, Computer Hardware, Electronics

An AI hiring system can treat every applicant the same and still leave some people feeling targeted. Researchers found that rejected candidates judged an automated interview differently depending on the race and gender of the avatar delivering the result.

Around 220 participants completed a simulated interview for a fictional customer support role with one of four photorealistic AI avatars. Everyone was rejected, yet perceptions of fairness shifted with the interviewer’s appearance. An algorithm audit could miss that reaction because candidates don’t experience the system as raw code. They experience a face asking questions and judging their answers.

Read more