Developer Backlash Hits Anthropic Over Claude

Anthropic is burning through developer goodwill to fund capital-intensive model training, creating a massive opportunity for open-source models to capture the developer market.

* **Eroding Trust:** Restricting subscription usage, charging extra fees for first-party CLI options, and using heuristic tricks to detect third-party wrappers are standard enshittification plays that alienate early adopters.

* **Open-Source Catching Up:** The capabilities gap between proprietary models like Claude Sonnet and open-source models like Qwen and GLM has shrunk significantly, making it viable to leave Anthropic's ecosystem.

* **Paradigm Shift:** The fragile and expensive "vibe coding" hype is giving way to pragmatic, developer-in-the-loop autocomplete workflows, reducing reliance on expensive, proprietary agent orchestration.