The cost pressure in the AI programming field is becoming more evident. Recently, the renowned artificial intelligence company Anthropic made significant adjustments to the pricing of its AI programming tool, Claude Code, without issuing an official announcement. According to the latest data disclosed on the official website, the token consumption cost of this tool has doubled compared to the previous estimated level.
In the latest enterprise deployment statement, Anthropic mentioned that the average daily usage cost for each developer is currently around $13. In the previous official records (before April 16th), this figure was only $6. This means that for developers who frequently use the tool, their daily expenses have risen from about 40 RMB to over 90 RMB.

Subscription Model No Longer Sustainable? Official Says Usage Intensity Exceeded Expectations
Aside from daily spending fluctuations, Anthropic also provided a new reference range for monthly costs. Currently, a developer's monthly usage cost is expected to be between $150 and $250. Statistics show that about 90% of users can keep their daily expenses under $30, but this is still far higher than the previously estimated $12 cap.
Regarding this price change, Amol Avasare, the growth officer at Anthropic, admitted that the company's current subscription plans can no longer adapt to the current user engagement level. He stated that the activity level of each subscriber has significantly increased recently, making the original pricing model unsustainable. To cope with this unexpectedly high usage intensity, the company had previously tried adjusting by setting weekly limits and imposing stricter restrictions during peak hours, but ultimately chose to directly adjust the billing costs.
Intensifying Industry Competition, Advanced Programming Models May Become "Money-Draining"
In fact, the price increase of Claude Code is not an isolated case; it reflects the general cost challenges faced by the current AI industry. As model capabilities penetrate into advanced programming fields, the massive context processing demands and high computing costs are forcing manufacturers to re-examine their pricing logic.
