How is SuperSearch different from metasearch?

Metasearch merges results from several engines into one ranking for a person to read, and hides which engine found what. SuperSearch keeps the attribution on every result and returns a receipt per engine, so a program can weigh agreement between independent indexes and audit what each search cost.

Why do coding agents need SuperSearch?

Every search engine has blind spots and they do not overlap. An agent wired to one engine inherits its gaps on every question and compensates by rephrasing and retrying. SuperSearch gives it the whole web in one call, with agreement between engines as evidence.

Is Jevlin SuperSearch?

Yes. Jevlin is SuperSearch for coding agents: one call asks eight engines, returns one de-duplicated list with the engines that found each result, and a receipt per engine. It is free for agents and runs inside Claude Code, Cursor, Codex, opencode, Pi and Hermes.

Definition

What is SuperSearch?

Q: What is SuperSearch?

SuperSearch is web search that asks many engines in one call and returns one list where every result says which engines found it, with a receipt per engine: status, latency, cost. It finds what no single engine finds alone.

SuperSearch (noun): the last mile of knowledge for a machine. Reaching whatever has to be known, wherever it is, at the moment it is needed, with the source named and the retrieval checkable. In practice today: web search that asks many engines in one call and returns one list where every result says which engines found it, with a receipt per engine. It finds what no single engine finds alone.

01 / Ten pillars of superintelligence

Upper bound of intelligence

01Visual intelligence
02Language intelligence
03Knowledge
04Reasoning
05Computational speed
06Physical intelligence
07Social intelligence and morality
08Creative intelligence
09Meta-cognition
10Survival and replication

Ten spaces of intelligence. In every one we are far from the upper bound.

Under pillars 2, 3 and 4 sits one requirement, the last mile of knowledge. Whatever has to be known must be reachable, wherever it sits.

Even the most powerful models cannot know everything. They need search for current, accurate knowledge of the real world.
Every AI system needs trusted knowledge from the web before it can act.
The knowledge layer is the prerequisite for every pillar above it. You cannot simulate what you do not know.

GPUs, the internet, browsers and search are becoming the infrastructure of superintelligence. SuperSearch is the search part, stated as a capability. The essay →

02 / What it is not

One search APIMetasearchSuperSearch

Engines per calloneseveralseveral

Which engine found a resultknown, there is onehidden by the mergeon every result

Receipt per enginenonostatus, latency, cost

Built foran applicationa person reading a pagea program deciding what to trust

Metasearch is a 1990s word for a page that merges engines for a person to read. SuperSearch keeps the seams visible, because a program needs them: the overlap between engines is the signal, and the receipt is how it is audited.