DEV Community

Cover image for Anthropic Almost Removed Claude Code From the Pro Plan: What Happened and What It Means for Indie Hackers
DevToolsPicks
DevToolsPicks

Posted on • Originally published at devtoolpicks.com

Anthropic Almost Removed Claude Code From the Pro Plan: What Happened and What It Means for Indie Hackers

Originally published at devtoolpicks.com


On April 21, 2026, developers opened Anthropic's pricing page and noticed something missing. Claude Code had a red X next to the Pro plan column. No email. No changelog. No announcement anywhere.

By the next morning, after Reddit, Hacker News, and X lit up with complaints, Anthropic reversed the change. The pricing page now shows Claude Code back on Pro. But the story does not end there, because what Anthropic's head of growth said while explaining the reversal is the part worth paying attention to.

Here is the full timeline, what it actually means right now, and what indie hackers using Claude Code should be thinking about.

What happened, in order

April 4, 2026. Anthropic blocked third-party agent frameworks, including OpenClaw and similar tools, from running on Pro and Max subscription plans. Anyone using these tools to run automated Claude workflows had to switch to pay-as-you-go API billing. For heavy users, the cost increase was reported at up to 50 times their previous monthly spend. No announcement.

April 21, 2026. Developers comparing the live pricing page against an archived version from April 10 noticed the change: Claude Code had been removed from the Pro tier. The support docs title changed from "Using Claude Code with your Pro or Max plan" to "Using Claude Code with your Max plan." Red X in the Pro column. Still no announcement.

April 22, 2026. After the backlash spread, Anthropic's head of growth Amol Avasare posted on X. He said it was "a small test on approximately 2% of new prosumer signups" and that existing Pro and Max subscribers were not affected. He then also said this: "Engagement per subscriber is way up. We've made small adjustments along the way (weekly caps, tighter limits at peak), but usage has changed a lot and our current plans weren't built for this."

Later that same day, he posted again: "Getting lots of questions on why the landing page and docs were updated if only 2% of new signups were affected. This was understandably confusing for the 98% of folks not part of the experiment, and we've reverted both the landing page and docs changes."

As of April 23, 2026. Claude Code shows as included in the Pro plan again on the pricing page. The test is apparently still running for 2% of new signups, but it is no longer visible to the rest of the world.

The current pricing (what is confirmed right now)

Plan Price Claude Code
Pro $20/month Yes (restored)
Max 5x $100/month Yes
Max 20x $200/month Yes

Claude Code is back on Pro for now. Existing subscribers were not affected during the test. That is the current state.

Why this is not just a "pricing page glitch"

The rollback was real, but the reason it happened is still sitting there.

Avasare said something specific: when Max launched a year ago, Claude Code was not part of it. Cowork did not exist. "Long-running async agents weren't a thing." The plans were designed for chat usage. That is not what people are actually doing now.

Since Claude Opus 4.7 launched in mid-April, users report Claude Code sessions running significantly longer than on the previous version, often with automated workflows running in the background for hours. The economics of a flat-rate $20 plan were not built for that kind of usage.

This is not a problem unique to Anthropic. GitHub recently suspended new signups for its flat-rate Copilot plan after similar compute issues. The underlying tension is the same everywhere: AI subscriptions were priced for people asking questions. Now they are being used to run agents for hours at a time.

Anthropic will not ignore this math forever. The April 21 test was probably the first of more to come.

What this means for indie hackers using Claude Code right now

If you are an existing Pro subscriber: Nothing changed for you during the test, and nothing has changed now. Your $20/month plan still includes Claude Code. If Anthropic decides to move ahead with restricting Claude Code on Pro, they would need to communicate that directly to existing subscribers. The backlash from this incident will make them more careful about how they do that.

If you are a new subscriber or considering signing up: Claude Code is listed as included in Pro right now. That is accurate as of today. But the honest answer is that this is a plan feature Anthropic is actively evaluating. If using Claude Code daily is core to how you work, Max 5x at $100/month gives you confirmed access and five times the usage limits.

If you are on Pro and heavy on Claude Code: This is the group most exposed to a future change. Avasare's comment about usage and plans "not being built for this" was a signal that more adjustments are coming. The test was paused, not cancelled.

The actual cost difference: An indie developer on Pro pays $240 a year. Moving to Max 5x is $1,200 a year. That is a $960 jump. Worth thinking about now rather than when the change gets forced.

The communication problem is the real issue

The feature itself is debatable. Plans getting restructured as usage changes is a normal business decision.

What is harder to defend is how this happened. No email to subscribers. No changelog. A pricing page change that contradicts a 2% test claim because the whole internet can see it. A support doc update that also contradicts the claim. Then a reversal with no direct apology to the people who were alarmed.

Simon Willison, who writes one of the most detailed Claude analysis blogs around, put it well: even if the test never goes live, the trust damage from this afternoon already happened.

For indie hackers who have built workflows around Claude Code, this incident is a useful reminder that any single tool as a dependency is a risk. That is not an argument against using Claude Code. It is still the best agentic coding tool available for most solo developers. But having a rough alternative in mind is not paranoid.

The Claude Code desktop redesign and parallel sessions update from April shows that Anthropic is still investing heavily in the product. And the Pro vs Max vs Cursor comparison covers how to think about which subscription actually makes sense for different usage levels. The basic answer right now: Pro is fine if you use Claude Code occasionally. If you are running it for hours daily, Max 5x is the more stable choice regardless of this incident.

For the broader context on how to get the most out of Claude Code as it stands today, the Claude Code solo developer guide is still current and worth reading alongside this.

FAQ

Is Claude Code still included in the Pro plan?

Yes, as of April 23, 2026, Claude Code is listed as included in the Pro plan on Anthropic's pricing page. The brief removal was reversed after developer backlash on April 22.

Were existing Pro subscribers affected?

No. Anthropic confirmed that existing Pro and Max subscribers were not affected during the test. The change, if it applies to anyone, is for a small percentage of new signups.

What is the difference between Pro and Max for Claude Code?

Both plans include Claude Code. Pro is $20/month with standard usage limits that reset every 5 to 8 hours. Max 5x is $100/month with five times the usage allowance and weekly limits rather than session-based ones. Max 20x is $200/month with 20 times the Pro limits.

Should I upgrade to Max now?

Not urgently, but worth considering if Claude Code is central to your work. Pro is fine for occasional use. If you run long automated sessions daily, the Max 5x limits are more predictable and the plan is less likely to be restructured.

What happened with third-party agent frameworks?

On April 4, Anthropic blocked tools like OpenClaw from running through Pro and Max subscriptions. Anyone using those tools now needs a separate pay-as-you-go API key. That change was not reversed.

The bottom line

Claude Code is still on Pro. The test was a test, not a final decision, and the backlash made Anthropic back off faster than they might have otherwise.

But the underlying economics they described are real, and they are not going away. Flat-rate plans were not designed for multi-hour agentic sessions, and every AI company is working through that same math right now.

Watch the next few months. If Anthropic announces a plan restructuring with proper notice and clear options for existing subscribers, that is fine. If they try another quiet change, expect the same backlash again, and faster this time.

For now: use Claude Code on Pro if that is your plan. Keep an eye on your usage. And if you are building something that depends on it daily, it is worth knowing what the Max 5x upgrade would cost before you are forced to find out in a hurry.

Top comments (0)