# Anthropic Releases Claude Sonnet 5 with Near-Opus Agentic Performance at Mid-Tier Cost

> The June 30, 2026 launch equips the Sonnet line with advanced planning and tool-use features previously limited to higher-end models while setting the new version as the default across Free and Pro plans.

*Published 2026-06-30 · By Marcus Vance*

Claude Sonnet 5 is Anthropic's most agentic Sonnet-class model to date, offering planning, tool use, and autonomous capabilities that approach Opus 4.8 performance at lower cost.

Anthropic released Claude Sonnet 5 on June 30, 2026.

The release targets the gap between mid-tier Sonnet models and higher-end Opus offerings in agentic tasks.

Claude Sonnet 5 builds directly on Sonnet 4.6 with targeted gains in planning and execution.

## What new agentic capabilities appear in Claude Sonnet 5?

Claude Sonnet 5 includes strong planning functions that allow step-by-step task decomposition.

Tool use now covers browser navigation and terminal command execution for real-world operations.

Autonomous capabilities enable the model to complete multi-step workflows with reduced oversight.

These elements together reach performance levels that previously required Opus-class models.

## Which benchmarks show the largest gains for Claude Sonnet 5?

Substantial improvement appears on the BrowseComp benchmark relative to Sonnet 4.6.

OSWorld-Verified scores also rise with the introduction of Claude Sonnet 5.

Both benchmarks evaluate practical agentic scenarios involving web interaction and operating-system control.

## How does availability and default access change with Claude Sonnet 5?

Claude Sonnet 5 serves as the default model for Free and Pro plans.

Max, Team, and Enterprise users gain access alongside the new default settings.

Claude Code and the broader Claude Platform and API also route requests to Sonnet 5 by default.

## What pricing applies during the introductory period for Claude Sonnet 5?

Introductory rates stand at $2 per million input tokens and $10 per million output tokens.

The promotional window extends through August 31, 2026.

Standard pricing after the period moves to $3 per million input tokens and $15 per million output tokens.

## What context window size does Claude Sonnet 5 support?

A 1M context window variant accompanies the standard release of Claude Sonnet 5.

The expanded window supports processing of lengthy documents and extended conversation histories.

## What do safety evaluations indicate about Claude Sonnet 5?

Overall rates of undesirable behaviors fall below those recorded for Sonnet 4.6.

Cybersecurity capabilities remain lower than those measured in Opus models.

## How does the release affect market positioning and stakeholder access?

Lower-cost access to advanced agentic features broadens reach to individual developers and smaller organizations.

Default inclusion on Free and Pro plans increases exposure without requiring plan upgrades.

API availability allows direct integration into production applications and agent frameworks.

## What reactions have followed the Claude Sonnet 5 announcement?

> Claude Sonnet 5 is built to be the most agentic Sonnet model yet. It can make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months ago, required larger and more expensive models. ... From today, Claude Sonnet 5 is available across all plans: it is the default model for Free and Pro plans...Anthropic

The system card describes Claude Sonnet 5 as an upgrade to Sonnet 4.6 with gains in agentic performance.

The same document notes that Sonnet 5 remains the most capable model within the Sonnet class.

It explicitly states that the release does not advance the capability frontier beyond Opus or Mythos models.

## What timeline elements mark the Claude Sonnet 5 rollout?

- Release date of June 30, 2026
- Introductory pricing window through August 31, 2026
- Default activation across Free, Pro, and API surfaces on launch day
- 1M context window variant available at release
- Safety evaluations published alongside the model card

## How do Claude Sonnet 5 specifications compare with prior models?

ModelInput Price (intro)Output Price (intro)Context WindowPrimary FocusClaude Sonnet 4.6Standard ratesStandard ratesStandardGeneral tasksClaude Sonnet 5$2 per million$10 per million1MAgentic planning and toolsClaude Opus 4.8Higher tierHigher tierStandardFrontier capabilities

## What comes next after the Claude Sonnet 5 launch?

Anthropic positions Sonnet 5 as the strongest model in its Sonnet family without crossing into Opus territory.

Subsequent iterations may refine the agentic features introduced in this release.

Users can expect continued emphasis on tool integration and autonomous execution in future updates.

The model card indicates no immediate plans to merge Sonnet and Opus capability boundaries.

Enterprise and developer feedback will likely guide refinements to planning and tool-use accuracy.

## Sources

1. [Claude Sonnet 5 is built to be the most agentic Sonnet model yet with plans, tool use, and autonomous capabilities at lower cost.](https://www.anthropic.com/news/claude-sonnet-5)
2. [Sonnet 5 upgrades Sonnet 4.6 with agentic performance gains but does not advance the capability frontier beyond Opus models.](https://www-cdn.anthropic.com/480e0bb54327b9622282e9c39a83a4f490ed377e/Claude%20Sonnet%205%20System%20Card.pdf)
3. [Claude Sonnet 5 reached stable release on June 30, 2026 with $2/$10 per million token promotional pricing and 1M context availability.](https://en.wikipedia.org/wiki/Claude_(AI))

---
Source: https://aiintelreport.com/frontier-models/anthropic-claude-sonnet-5-agentic-release
Index: https://aiintelreport.com/llms.txt · Full text: https://aiintelreport.com/llms-full.txt