Anthropic Is Now Paying Subscribers to Stay as Rivals Close the Gap
When a company starts handing out money to the people who already pay it, it is usually worried about something. Anthropic has just done exactly that, and the timing says more than the press release does.
- Anthropic now adds monthly API credits to eligible Claude plans, starting at $100 a month for Max 5x and $200 for Max 20x, and the credits last only while the plan is active
- The credits cover the Claude API, Claude Managed Agents, the Claude Agent SDK and the Playground, and they do not cover interactive Claude Code or the Claude apps
- Subscribers can add a card for future credit purchases, and nothing is charged today
- Claude was once the clear front runner, but DeepSeek, Alibaba, Z.ai's GLM models, European startups and OpenAI have all closed the gap, often at lower prices
- Meta's Muse overtook ChatGPT, Claude and Grok on the App Store in September, which keeps the pressure on every paid assistant
- The credits read as a retention move, because they are only worth using if you stay subscribed
What Anthropic is actually giving away
Anthropic announced on October 7 that Max and Team plans now come with monthly API credits. On Max 5x the credit is $100 a month, and on Max 20x it is $200, according to early coverage of the announcement. Team seats draw from a shared pool worth up to $500 a month. The credits apply to the Claude API, Claude Managed Agents, the Claude Agent SDK and the Playground, so they are aimed at people building with Claude, not at people chatting with it.
There are limits worth knowing. Interactive Claude Code in the terminal, the IDE, the desktop app and the web is not covered, and neither is extra usage in the Claude apps. Free, Pro and Enterprise plans are not eligible, and the credit refreshes each month while the plan stays active. Subscribers can also add a card for future credit purchases, and nothing is charged today. Rollout is staged, so some accounts may not see the offer yet.
Why hand out credits now
A year ago, the question of which AI to build on had an obvious answer for many developers, and that answer was Claude. That is no longer true. DeepSeek, Alibaba's Qwen models and Z.ai's GLM family have all reached a level that many teams treat as good enough, and they often cost far less to run. European AI startups have caught up in their own corners, and OpenAI has stayed close on the things that matter to most buyers. Open-weight models make the picture harder for everyone selling a closed one, because a team can host them itself and stop paying per token.
Then there is Meta. Its Muse app climbed to the top of the US App Store in September, passing ChatGPT, Claude and Grok on the free chart, according to Sensor Tower data reported at the time. Rankings move quickly and Muse may not hold the top spot, but the signal is clear. Attention in consumer AI is up for grabs, and the gap between the leaders is much smaller than it used to be.
A proposed class action filed against Anthropic in the Northern District of California claims Claude Max subscribers pay for 5x or 20x the usage of the base Pro plan but receive far less. Plaintiff Karl Kahn says internal Anthropic emails and his own account activity point to real multipliers closer to 3.5x and 6-8x, not the numbers in the... - why anthropic claude max plan delivers less usage - class action lawsuit anthropic false advertising claude max
A credit that only works if you stay
Look at how the offer is built. The credit is monthly, it expires with the billing cycle and it only exists while the plan is active. Nothing about it helps you if you cancel. It also points customers toward the API, Managed Agents and the Agent SDK, the products where developers form habits and where switching later is painful. That is a sensible way to make your best customers a little more attached.
This is not a criticism. Retention offers are normal in a market where products are becoming alike, and a developer who gets $100 or $200 of extra API usage each month has a real reason to keep an eye on Claude before trying something cheaper. But a company that was comfortably ahead has no need to do this. The move says Anthropic knows that being first is no longer enough.
What it means if you build with AI
For founders, the useful reading is practical. If you already pay for Max and build on the API, the credit is extra usage you do not pay for and worth claiming once your account shows it. If you do not, treat the offer as a sign of where pricing is heading, because every serious lab is now competing on cost and loyalty as much as on benchmark scores. Test two or three models on your own work before you commit, and keep your code easy to move.
The race is now close enough that the best answer can change from one quarter to the next. Anthropic is paying to stay in front of its customers. That tells you the customers have options.
This article is posted in AI News, check it out for more related stories.
A federal class-action lawsuit accuses Anthropic of misrepresenting how much usage its $100 Max 5x and $200 Max 20x Claude plans actually deliver. Court filings show Max 20x provides only about 1.7 times the weekly Sonnet hours of Max 5x, despite costing double, and a hearing on Anthropic's motion to dismiss is set for November 6, 2026. - Claude Max 20x usage limits lawsuit against Anthropic - does Claude Max 20x deliver promised token capacity
