Three different things people conflate
When founders say "I want to get money from Anthropic," they usually mean one of three very different things, with very different odds and processes. Conflating them wastes time. They are: (1) API credits, (2) startup support programs, and (3) the Anthology Fund venture investment.
1. API credits — the easy, likely "yes"
The most accessible value is Claude API credits for startups — credits that offset your model costs as you build. This is not cash and not equity; it lowers your burn. It is the lowest-effort, highest-odds path and the right first ask for almost any startup building on Claude.
If your product calls Claude (an agent, a triage layer, anything), apply for startup credits early. It is close to free money in the literal sense that it reduces a bill you would otherwise pay.
2. The Anthology Fund — venture investment, not a grant
The Anthology Fund (associated with Anthropic and Menlo Ventures) makes equity investments in AI-ecosystem startups, with AI-payments infrastructure explicitly in scope. Crucial distinction: this is venture capital. There is no application form that returns a check. You get in the way you get into any fund — a crisp pitch, traction or an exceptional wedge, and usually a warm introduction.
For an agent-economy or agent-payments startup, the credible signals are a working product, real (even if small) external demand, and a defensible wedge — not a deck full of projections.
3. Research grants — academic, intermittent, narrow
Anthropic's Economic Futures program has offered research awards ($10k–$50k) for empirical economics research on AI's impact. These are for economists studying labor markets and productivity — not for building infrastructure — and they open and close intermittently. If you have a genuine empirical dataset about AI's economic effects, it can fit; if you are building a product, it does not.
What to actually do
Sequence it: apply for API credits now (easy win). Build something real with genuine usage. Then — once you have a working product and early demand — pursue the Anthology Fund through a warm intro, ideally via visible presence in the Claude/MCP ecosystem. Do not wait for a grant that does not exist for product builders.