A developer's month-long wait for Anthropic support exposes a growing gap in how AI companies handle customers who aren't enterprise accounts.
Nick Vecchioni waited over thirty days for Anthropic to respond to a support request. That sentence alone wouldn't be remarkable if Anthropic weren't one of the most valuable AI startups on the planet, recently closing a $2 billion funding round at a $60 billion valuation. Yet a blog post documenting that exact experience gained traction on Hacker News this week, striking a nerve with developers and small teams who rely on Claude's API but feel invisible when something breaks.
The timing is uncomfortable for Anthropic. The company has spent the last year positioning itself as the safety-conscious, enterprise-ready alternative to OpenAI. It landed partnerships with Amazon Web Services, expanded Claude's capabilities into agentic tool use, and pushed aggressively into the enterprise market. But enterprise polish means little if a paying developer cannot get a human being to respond to a support ticket within a month.
Vecchioni's experience is not unique to Anthropic. It reflects a structural issue across the AI sector. Companies building foundation models are pouring billions into compute, talent, and research. Support infrastructure is treated as an afterthought, something to scale later once the product matures. The result is a two-tier system. Large enterprise clients get dedicated account managers and Slack channels. Everyone else gets a ticket queue that may or may not surface a response.
OpenAI faced similar criticism throughout 2023 and 2024, particularly from developers using its API who encountered billing errors or sudden account suspensions with no clear recourse. Google DeepMind, while less dependent on individual developer revenue, has also struggled with response times for Vertex AI issues raised by smaller customers. As Forbes recently pointed out, the pattern is consistent across the sector: rapid product launches followed by slow, underfunded customer operations.
The economics behind this are straightforward even if the consequence is frustrating. A single support engineer costs between $80,000 and $150,000 annually in total compensation, depending on location and seniority. When a startup is burning through hundreds of millions on NVIDIA GPUs and researcher salaries, adding fifty support headcount barely registers on the spreadsheet. But those fifty people could meaningfully reduce response times for thousands of paying customers.
Why Developers Matter More Than AI Companies Think
Here is the part that often gets lost in boardroom conversations. Developers and small teams are the ones building applications, integrations, and workflows on top of these models. They are not the largest revenue source today, but they function as a distribution network. A solo developer building a tool that reaches 10,000 users is creating downstream adoption that no marketing campaign can replicate. When that developer hits a wall and cannot get support, they switch providers. The switching cost for swapping one API key for another is remarkably low.
Anthropic charges for Claude access through usage-based pricing. Developers paying even a few hundred dollars a month are real customers, not free riders. Expecting them to wait weeks for a response erodes trust in a market where trust is already fragile. The AI sector is littered with companies that overpromised and underdelivered. Support is one of the few tangible signals a customer can use to judge whether a company will be reliable over the long term.
There is also a competitive dynamic at play. Mistral, Cohere, and open-source alternatives like Meta's Llama ecosystem are actively courting developers who feel underserved by the largest providers. A poor support experience does not just lose one customer. It creates a narrative that competitors will happily amplify.
The Hacker News thread on Vecchioni's post may have had zero comments, likely because the post gained only nine upvotes, but the broader signal is clear. When someone documents a month-long wait for basic support from a $60 billion company, it resonates because it confirms what many developers quietly suspect. The AI industry is building for enterprises and treating everyone else as an afterthought.
Anthropic and its peers need to recognize that the developer ecosystem is not a sideshow. It is the foundation of long-term platform dominance. Underinvesting in support is a short-term cost-saving measure that carries a long-term strategic risk. The companies that figure this out first will retain the builders who ultimately decide which models end up inside the products the rest of the world uses.
