What’s New with Sonnet 5?
Claude Sonnet 5 is Anthropic’s latest mid-tier model, but don’t let the “Sonnet” label fool you — it nearly catches up to Opus 4.8 on agentic performance while costing significantly less.
| Spec | Claude Sonnet 5 | Claude Opus 4.8 |
|---|---|---|
| Input Price (after Sep 1) | $3 / million tokens | $4 / million tokens |
| Output Price (after Sep 1) | $15 / million tokens | $25 / million tokens |
| Intro Pricing (through Aug 31) | $2 / $10 per million tokens | — |
| Availability | Free, Pro, Max, Team, Enterprise, API | API, Enterprise |
| Focus | Agentic AI, coding, tool use | High-capability tasks |
Built for Agentic Workflows
The big differentiator? Sonnet 5 is designed for agentic AI — where models make plans, use tools, browse the web, run terminals, and execute multi-step tasks autonomously.
Early access testers reported that Sonnet 5:
- Finishes complex tasks where previous models would stall halfway
- Self-checks its output without being prompted
- Handles multi-step software engineering — writing tests, implementing fixes, and verifying results in a single pass
One developer noted: “I asked Sonnet 5 to investigate a bug. Unprompted, it wrote a reproducing test, implemented the fix, then stashed it to confirm the bug came back without the change. All in a single pass.”
Pricing That Makes Sense
Starting at $2 per million input tokens and $10 per million output tokens through August 31 (going to $3/$15 after that), Sonnet 5 offers Opus-competitive agentic performance at a fraction of the cost. For heavy enterprise users running agent workflows, this could mean serious savings.
Safety Improvements
Anthropic’s safety assessments show Sonnet 5 has a lower overall rate of undesirable behaviors than Sonnet 4.6, and it’s safer in agentic contexts. It also scores lower on cybersecurity risk compared to Opus models — a deliberate tradeoff for broader deployment.
Who Is This For?
- Developers building AI-powered coding assistants
- Enterprises running automated agent workflows
- Power users on Claude Pro / Max who want faster, cheaper responses
- Anyone tired of burning through token budgets on agentic tasks
FAQ
Q: When is Claude Sonnet 5 available?
A: It’s available now — June 30, 2026. It’s the default model for Free and Pro plans starting today.
Q: How does Sonnet 5 compare to Opus 4.8?
A: Sonnet 5 delivers agentic performance close to Opus 4.8 at roughly 40% lower cost. For many agentic tasks, the gap is negligible.
Q: Can I use Sonnet 5 via API?
A: Yes. Developers can access claude-sonnet-5 via the Claude API and Claude Code.
Q: Will Sonnet 5 replace Opus?
A: Not entirely. Opus still outperforms on complex reasoning and cybersecurity tasks. But for most agentic workflows, Sonnet 5 is the smarter choice.
Final Verdict
Claude Sonnet 5 is a smart release from Anthropic — better agentic performance, lower cost, and safer by design. If you’re running AI agents or just tired of expensive token bills, this is the model to switch to.
Score: 8.5/10
👉 Try Claude Sonnet 5 on Anthropic’s website
This post contains affiliate links. As an Amazon Associate, PC Master Deals earns from qualifying purchases.