If you use Claude on the Free or Pro plan, something changed on your account this week and you didn't have to do anything to get it.
On June 30, 2026, Anthropic announced Claude Sonnet 5 and made it the default model for Free and Pro plans. In Anthropic's words, "it is the default model for Free and Pro plans." No upgrade button, no settings toggle, no new bill. If you opened Claude today, you were already using it.
For a small-business owner who feels a step behind on all this, that's the part worth knowing. AI that can plan a task and carry it through every step used to sit behind bigger, pricier models. Now it's the everyday model on the tiers most people are already on.
What "more agentic" means when you run a business
Anthropic describes Sonnet 5 as its most capable Sonnet model for agent-style work. Here's how they put it:
"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."
Strip out the technical words and it comes down to this. An agentic model tries to do the whole job. You ask it to research three suppliers and put them in a table, and it goes and does the steps rather than telling you how you might do them.
That's a big deal for someone wearing every hat. Most owners don't need a chatbot that gives good advice. They need something that can take a multi-step task off the list. Sonnet 5 is built to attempt more of that kind of work, and it's the model already sitting in your account.
Anthropic also says Sonnet 5 is a "substantial improvement over its predecessor, Sonnet 4.6, on important aspects of agentic performance like reasoning, tool use, coding, and knowledge work." If you tried an earlier version of Claude a few months back and it felt close-but-not-quite on a task, that's a fair reason to try the same task again.
The price story, for anyone who wants to build something small
Most owners will use Claude straight from the app and never think about pricing. But if you or someone you hire wants to wire up a small automation, the cost of running it also came down.
Through the app, Sonnet 5 is the default on Free and Pro. Through the Claude API (the pipe developers use to connect Claude to your own tools), the model runs under the name claude-sonnet-5, and Anthropic set the price like this:
- Now through August 31, 2026 (introductory): $2 per million input tokens and $10 per million output tokens.
- After August 31, 2026 (standard): $3 per million input tokens and $15 per million output tokens.
Introductory API price per million input and output tokens for claude-sonnet-5, in effect through August 31, 2026.
Tokens are just chunks of text, roughly a few characters each. A million of them is a lot of back-and-forth. The exact math is less useful than the direction of it. Anthropic priced this model below its top-tier model, Opus 4.8, while saying Sonnet 5's performance is close to it. More capable, and cheaper to run at scale than the model it's compared against.
If a custom automation was on your someday list because it felt expensive to run, the ground under that decision moved a little in your favor.
Why this fits the way proficiency grows
At LaunchReady we teach The 7 Levels of AI Proficiency, and the honest truth is that the tool getting better does not move you up the levels by itself. The early levels are about using AI at all and learning to ask it good questions. The higher levels are about giving it context from your business and designing a workflow it can carry.
A more capable default model helps most if you meet it with a real task. You already have the tool, so the useful response to this news is to point today's Claude at one job you keep putting off and see how far it gets. That's how proficiency climbs, one rep at a time.
What to try this week
Pick one task you dread and hand it over. A few that work well for owners:
- Turn a mess into a draft. Paste in your rough meeting notes and ask Claude to write the follow-up email and a short list of next steps.
- Do the boring research. Ask it to compare three tools or vendors you're weighing and lay the differences out in a table.
- Write the thing you never get to. A job posting, a set of FAQ answers for your site, a first pass at a policy.
You'll learn more from one attempt than from reading three more articles about AI.
Your next step
Open the Claude you already have and give it one task off your list today. If you want a clearer picture of where you and your team stand and what to build next, the 7 Levels of AI Proficiency assessment is a short, honest read on your starting point.
Related reading: Level 4: The Commander (Context Engineer).
Sources
Frequently Asked Questions
Do I need to change anything to get Claude Sonnet 5?
No. Anthropic made it the default model for Free and Pro plans, so if you use Claude on either tier you already have it.
Is this only useful if I know how to code?
No. The app upgrade helps anyone who uses Claude for everyday work. The API pricing is the part to watch if you or a developer want to build a custom automation.
How does it compare to Anthropic's top model?
Anthropic says Sonnet 5's performance is close to that of Opus 4.8, its top-tier model, but at lower prices. It also calls Sonnet 5 a substantial improvement over the earlier Sonnet 4.6.
How much agent-style work can I run on the free plan?
Anthropic's announcement didn't spell out usage limits for the Free tier, so that's an open question. If you're leaning on Claude for heavier daily work, the Pro plan gives you more room.
Find your AI Proficiency level
The free 7 Levels assessment places you across seven stages of AI capability. Under ten minutes. Research-backed scoring.