Claude Sonnet 5 Is Now the Default Model in Claude Code
Claude CodeAnthropic released Claude Sonnet 5, now the default model in Claude Code, featuring a native 1-million-token context window and introductory pricing of $2/$10 per million input/output tokens through August 31, 2026 (standard pricing is $3/$15 per Mtok). Anthropic positions Sonnet 5 as approaching Opus 4.8-level performance on agentic coding, reasoning, and tool-use benchmarks at a fraction of the cost. The model is available across all Claude plans and the Claude API under the identifier claude-sonnet-5, and ships with real-time cyber safeguards enabled by default. Early third-party testing shows mixed but substantial reactions, with some reviewers calling it one of the strongest coding models yet and others flagging inconsistent results compared to the hype.
Featured Video
A video we selected to help illustrate this changelog
Key Takeaways
- A 1-million-token context window is now native to the default Claude Code model, letting sessions hold much larger codebases and longer histories before compaction kicks in.
- Promotional pricing of $2/$10 per Mtok runs through August 31, 2026, before reverting to standard $3/$15 per Mtok: a meaningful window for cost-sensitive teams to scale agentic usage.
- Sonnet 5 targets near-Opus performance at higher effort settings, per Anthropic's own benchmarks, aiming to make Opus-tier reasoning accessible at Sonnet-tier pricing.
- Third-party reviews are split, with some outlets calling it the best coding model yet and others arguing the upgrade over Sonnet 4.6 is incremental relative to the hype.
- Real-time cyber safeguards are enabled by default, alongside a reported reduction in exploit-development capability, reflecting Anthropic's continued safety-first framing for coding-capable models.
- The release coincides with wider strategic positioning, with press coverage tying the aggressive pricing to Anthropic's push for broader agentic-AI adoption ahead of an anticipated IPO.
Sources & Mentions
5 external resources covering this update
NEW Claude Sonnet 5 vs Opus 4.8! (Full Review)
Youtube
Anthropic launches Claude Sonnet 5 as a cheaper way to run agents
TechCrunch
Anthropic launches Claude Sonnet 5 at a steep discount to its top model as the company races toward a blockbuster IPO
VentureBeat
Anthropic Sonnet 5: It closes the gap with Opus 4.8, and is cheap until August
The New Stack
Anthropic Claude Sonnet 5 vs Sonnet 4.6 vs Opus 4.8: Agentic Coding Benchmarks, API Pricing, and Cost-Performance Tradeoffs Compared
MarkTechPost
A New Default Model for Claude Code
Anthropic introduced Claude Sonnet 5 on June 30, 2026, and immediately made it the default model in Claude Code (available starting version 2.1.197). The headline upgrade for developers is a native 1-million-token context window, letting Claude Code hold much larger codebases, longer conversation histories, and more extensive tool output in a single session without aggressive compaction.
Pricing and Availability
Anthropic set introductory pricing at $2 per million input tokens and $10 per million output tokens, available through August 31, 2026. After the promotional window, standard pricing settles at $3/$15 per million tokens. Sonnet 5 is accessible everywhere Claude ships: Free, Pro, Max, Team, and Enterprise plans, the Claude API (model ID claude-sonnet-5), and the Claude Platform, in addition to Claude Code.
Performance and Positioning
Anthropic describes Sonnet 5 as its most capable Sonnet-tier model to date, closing much of the performance gap with Opus 4.8 on agentic tasks: multi-step planning, sustained coding and debugging sessions, browser and terminal tool use, and complex reasoning, while running at meaningfully lower cost. At higher effort settings, Anthropic says the model approaches Opus-class results on select benchmarks, positioning it as the practical default for everyday agentic coding work where Opus-level spend isn't justified.
Independent coverage has been split. Some reviewers praised it as the best coding model they've tested, citing strong results on real-world coding, browser automation, and game-development test suites. Others pushed back, arguing the jump from Sonnet 4.6 feels incremental for the price and that select benchmarks underperformed expectations. TechCrunch and VentureBeat both framed the release within Anthropic's broader business strategy: undercutting the cost of top-tier models to widen agentic-AI adoption ahead of a widely anticipated IPO.
Safety Posture
Anthropic reports lower rates of undesirable model behaviors compared to its predecessor, along with real-time cyber safeguards enabled by default and a substantially reduced capability for exploit development: details relevant to enterprises evaluating Claude Code for security-sensitive codebases.
What This Means for Claude Code Users
For everyday Claude Code users, the practical impact is threefold: sessions can hold far more context before hitting compaction, per-token costs drop significantly during the promotional window, and agentic workflows (subagents, background jobs, /goal, /loop) should see improved reliability on longer-horizon tasks now that the default model has a larger working memory and stronger tool-use performance.