Just a few months ago, Claude Code was widely seen as one of the most powerful and developer-friendly AI coding tools on the market. In 2025, it had become a favorite among power users, praised for its speed, model quality, and generous subscription plans.
But in early 2026, the tone has shifted dramatically. In a matter of days, enthusiasm turned into frustrationโand in some corners of the developer community, outright anger.
So what happened? And why is Claude Code suddenly being criticized by the very users who once championed it?
The GitHub issue that sparked the backlash
The controversy began quietly, with a simple question posted on a GitHub issue: โWhy does OpenCode no longer work with the Claude Max plan?โ
That single issue quickly escalated into a broader debate across GitHub, X (formerly Twitter), and developer forums. The reason behind the problem soon became clear: Anthropic had restricted the use of its Opus model when accessed through third-party tools.
In other words, developers who were paying for Claudeโs higher-tier subscriptions could no longer use those benefits inside external coding environments such as OpenCode, Cursor, or Windsurf.
How Claude subscriptions used to work?
Until recently, Claude Code subscriptions were unusually generous. Users could choose between Pro, Max, or Max Pro plans, with the Max planโpriced around โฌ180 per monthโoffering significantly higher usage limits.
That plan unlocked:
- Unlimited access to Claude chat
- Advanced research features
- Access to Anthropicโs best models, including Opus
- Claude Code in the terminal (via CLI)
Crucially, developers could also use that same subscription inside third-party tools. This made Claude Max especially attractive when paired with OpenCode, which had rapidly gained popularity for its fast, fluid coding experience.
Why OpenCode mattered so much to developers?
I’m floored Anthropic aggressively cut off paying customers from using Claude Max subscriptions with open source agents.
They’re speedrunning the journey from forgivable startup to loathsome corporation before any exit!@_catwu โ I’d strongly urge you to reconsider this. pic.twitter.com/C46QajE1LC
โ AJ Stuyvenberg (@astuyve) January 9, 2026
OpenCode had been โthe tool everyone was talking about.โ Many developers preferred its interface and workflow over Claudeโs own terminal-based experience.
For them, the ideal setup was simple: pay for Claude Max, then plug that power directly into OpenCode.
That combination delivered faster responses, smoother interactions, and what many described as a superior developer experience.
So when Anthropic suddenly blocked Opus usage outside of its own tools, the reaction was immediate.
โI just paidโand now itโs goneโ
The loudest complaints came from users who had subscribed to Claude Max specifically to use it with OpenCode. Some reported that they had upgraded their plan only days earlier.
Then, without warning, their workflow broke. OpenCode started returning errors, effectively telling them that this usage was no longer allowed.
From their perspective, the frustration was understandable. They felt they had paid for a capability that was abruptly taken away.
Workaroundsโand the risk of getting banned
As with most developer controversies, workarounds quickly appeared. Community members shared methods to continue using Claude Max through OpenCode by chaining tools or modifying requests.
But these solutions came with a major caveat: they violated Anthropicโs terms of service.
Several developers warned that using such hacks could lead to account suspension. For many, that risk simply wasnโt worth itโespecially at this price point.
The broader criticism: lock-in and ecosystem control
As the debate expanded, criticism shifted from the technical limitation itself to what it represented.
Many developers accused Anthropic of trying to force users into its own ecosystem: Claude chat, Claude Code, and its newly released workflow tools.
The concern wasnโt just about OpenCodeโit was about control. By limiting where its models can be used, Anthropic keeps pricing power and product direction firmly in its own hands.
The business reality behind the decision
From a business standpoint, the move is easier to understand.
Developers using Claude Max intensively can generate massive token usage. When routed through third-party tools, that usage becomes harder to controlโand more expensive to subsidize.
One example shared in the discussion highlighted the issue clearly: a developer using Claude purely via API reached โฌ200 in usage in just three days.
Compare that with a โฌ200 monthly subscription that allows near-unlimited usage, and the imbalance becomes obvious.
For power users, Claude Max was an incredible deal. For Anthropic, it may not have been sustainable.
The buffet analogyโand why it didnโt convince everyone
The pushback against Anthropic right now is pretty stupid.
Here's basically what happened:
1. They invent a new concept of an all-you-can-eat subscription to their amazing buffet restaurant.
2. It's the best food in town, so everyone wants it.
3. …— แด แดษดษชแดส แดษชแดssสแดส ๐ก๏ธ (@DanielMiessler) January 10, 2026
A viral post by Daniel Mler compared Claudeโs subscription model to an all-you-can-eat buffet: profitable because most customers donโt fully exploit it, but costly when a few do.
The analogy gained tractionโbut also heavy criticism. Many developers argued it oversimplified the situation and ignored the expectations set by the product itself.
After all, Claude Max was marketed as a premium, power-user-friendly plan.
Not all criticism is equal
As often happens online, once the backlash started, every possible complaint resurfaced.
Some criticized Claude Code for automatically adding itself as a co-author on Git commits. Others called the behavior โcringeโ or โobnoxious.โ
Yet this feature can be disabled in secondsโmaking these critiques feel more like emotional pile-ons than substantive issues.
The real debate remains the restriction itself, not cosmetic annoyances.
A calmer take: frustration, not betrayal
While many developers feel betrayed, not everyone shares that sentiment.
Some acknowledge the inconvenienceโand even the disappointmentโwithout seeing it as a moral failure. Tools change. Pricing models evolve. Businesses protect their margins.
From that perspective, Claude Code is still an exceptionally strong product. And if another tool eventually does better, switching remains an option.
What happens next?
The future of Claude Code will be shaped by familiar forces: competition, pricing pressure, developer trust, and market alternatives.
If Anthropic maintains its technical edge, many users will stayโlimitations and all. If better options emerge, loyalty will fade quickly.
In the end, this episode may be less about Claude Code itself and more about a broader truth in developer tooling: no platform is entitled to long-term affection.
Tools are tools. They are adopted, evaluated, replacedโand sometimes forgiven. The debate is far from over. And if the situation evolves, it will almost certainly reignite.
