Our mission and vision
Make web search so cheap that no one ever architects around it.
Something inverted in the economics of AI. Model prices fall every quarter, but the moment your agent needs to know something true about the world, a current fact, a real document, an actual opinion, the meter starts running. Search APIs charge per request, agents search in loops, and the products that search the most are punished the most. Builders respond the only way they can: they cache too aggressively, they cap their agents, they ship products that know less than they should.
We think that's backwards. Grounding an answer in real sources should be the default behavior of every AI product, not a budget line to be managed. So we built laurel.dev to make source-backed web context effectively free: one call returns ranked results and clean Markdown excerpts a model can cite, queries are routed to the sources that actually answer them: news to news engines, code to code hosts, papers to paper indexes. Aggressive caching means the same question never costs twice.
Unlimited search, flat price, no meter anxiety. Search shouldn't be the thing stopping you from building.